Public report — CAI, published 15 Sep 2026.
Concrete security findings (CVE IDs, secret matches, dependency versions) are hidden in this version;
ask the repo owner for the full report.
Public
Codebase surveyMeasured under the Code Assurance Index · rubric rubric-2026.09.11 (frozen) · verify this surveyFiledcd_5220ecd610be41db903e00d1e9942cdf
Filed 15 September 2026, 12:43 UTC
Public
Small · 8,601 LoC · 6 projects · rebuild ~2.6 person-years · weakest lens: Architecture (78%)
Findings by grade
15 critical40 serious37 minor1 could not be resolved — could be critical — see Limitations
This survey was produced by
Watchdog
Producer
Canine Development
Analyzer
Watchdog engine 1.0.0
Measured
15 September 2026, 12:36 UTC
A measurement, not a certificate. The Code Assurance Index does not certify,
approve or guarantee this codebase; it records a reproducible number and the evidence it was computed from. The
standard is authored by Canine Development, who also build Watchdog — its only implementation today. That is said
here so the number is checked rather than believed.
Grounded in facts. Every number here is computed, not narrated — reproducible, tool-backed, and traceable to a line of code. How to trust this ▸
58findings with an exact file:lineof 92 — the remainder are repo-wide signals (a dimension-level measurement, not a single line); open any file:line and verify
83/116dimensions across the health lenses8601 LoC · 6 projects — wide & deep
Preview (pre-1.0). This repo hasn't declared a stable release, so it's judged against a relaxed, pre-production bar.
The system holds a strong overall standing of 79%, indicating a healthy asset that is largely maintainable and ready for production. However, beneath this solid surface lies a critical ambiguity in how the code is organized, which poses a risk to long-term delivery speed and team scalability. While the current metrics suggest stability, the underlying structure requires clarification to ensure it supports future growth without hidden costs.
The value at stake involves a modest codebase of roughly 8,600 lines of production logic, with a rebuild effort estimated at 2.6 person-years or approximately €380,000. This represents a manageable investment, but the composition reveals that 57% of the code is branching logic, suggesting complex business rules that are not yet clearly separated from infrastructure concerns. If these boundaries remain blurred, every change risks unintended side effects, increasing the cost of maintenance and slowing down feature delivery.
The primary theme is structural ambiguity. Most code does not fit into standard architectural layers, not because the design is poor, but because naming conventions are unclear. This makes it difficult for new engineers to understand the system or for teams to predict where changes will ripple. Without clear layer definitions, the codebase appears fragmented, leading to potential defects and increased onboarding time. Clarifying these structures is the highest-leverage action, as it unlocks the ability to safely refactor and extend the system.
A secondary concern is security posture. While the overall readiness is good, the surface area for anonymous access needs deliberate review. Keeping open endpoints intentional and documented is crucial to prevent security exposure. Additionally, interfaces are currently too broad, forcing clients to depend on more than they need. Splitting these into focused roles will improve reliability and reduce coupling, making the system more resilient to change.
Genuinely good aspects include high maturity and accessibility scores, indicating that the team has established strong practices for testing and user experience. The system is observable and safe to operate in its current state. However, the picture is partial; key areas like domain modeling and event-driven patterns were not measured, so their impact remains unknown.
Focus first on naming and organizing the code layers. This single step provides the most leverage, transforming ambiguity into clarity and enabling all other improvements. Without this foundation, other optimizations will yield diminishing returns.
How the score is built — each lens's share of the headlineWidth is the lens's weight in the worst-heaviest fold (the weakest area pulls hardest); colour is that lens's own band. A lens fixes the score in proportion to its width.
1.3× (at 79% quality) — the last 20% of quality is most of the work
Size & shape
Small · effort split not classified for 130 line(s) outside the .NET model (the tier breakdown is a C#-only syntax walk)
This codebase represents roughly ~2.6 person-years of build effort (about ~€380,000 to rebuild). Its weakest lens is Architecture at 78% — the part of that asset most exposed by the findings below.
How we model this: boilerplate at a scaffolding rate + logic × domain Standard (×1.2) — service/app, high decision density × a 1.3× quality factor, at €60–95/h; indicative, ±~30%. Indicative only — most sensitive to the hourly rate and the domain tier (both tunable in config).
Top priorities
The highest-leverage moves; the full ranked list is in the Roadmap below.
1
Resolve the 9 High finding(s) in Static Analysis (SAST) — start with publish-packages.yml (4), NoiseStore.cs (2), RegistryStore.cs (2).
Of everything flagged, the best return on effort is: The domain core is a small share of production code, but most of the rest matched no layer vocabulary at all — so this is not yet an anemic-domain finding. The namespace/path convention could not place that code, which makes the composition above a statement about the naming, not about the design. Name the layers (or check that the repository's conventions differ from the ones this check knows) before reading a thin domain into it. The rest can wait behind it.
Evidence: priority ranking: top of 5 ranked by impact/effort
→ The domain core is a small share of production code, but most of the rest matched no layer vocabulary at all — so this is not yet an anemic-domain finding. The namespace/path convention could not place that code, which makes the composition above a statement about the naming, not about the design. Name the layers (or check that the repository's conventions differ from the ones this check knows) before reading a thin domain into it.
Architecture — module dependency graph
Project dependencies, layered top-to-bottom; arrows show direction. Any dashed red edge points upward or sideways — a layering smell or cycle. A clean layered graph has none.
Architecture — module dependency matrix
Rows and columns are the same modules, ordered so that a module only depends on ones above it. A cell means the row depends on the column, and its number is how many type pairs create that dependency. Read one thing: is anything above the diagonal? A mark there is a dependency cycle. (A cycle is all this shows — an unusual but cycle-free dependency sits below the diagonal like any other.)
11 modules, 5 dependencies. Every dependency points down the layering — no cycles.
Module dependency matrix. The row depends on the column; the number is how many type pairs create the dependency. A cell above the diagonal is part of a dependency cycle.
Findings mapped to OWASP categories; the specific CVEs/secrets are in the Security dimension cards below and findings.md (redacted only on the public version of this report).
OWASP category
Findings
Severity
A03:2021 — Injection
16
High / Critical
A02:2021 — Cryptographic Failures
7
High / Critical
A06:2021 — Vulnerable & Outdated Components
1
Medium
Roadmap
First, clarify the codebase structure by explicitly naming layers to resolve ambiguous composition, then split fat interfaces into focused roles to improve modularity. Next, audit anonymous endpoints to ensure open access is deliberate and documented, and propagate cancellation tokens through async methods for responsive behavior. Finally, resolve the nine high-priority static analysis findings, prioritizing issues in the pipeline configuration and core store classes.
Ranked by impact ÷ effort. "Helps" is the estimated gain on the 0–100 health score.
Do this
Helps
Effort
Dimension
Resolve the 9 High finding(s) in Static Analysis (SAST) — start with publish-packages.yml (4), NoiseStore.cs (2), RegistryStore.cs (2).
The domain core is a small share of production code, but most of the rest matched no layer vocabulary at all — so this is not yet an anemic-domain finding. The namespace/path convention could not place that code, which makes the composition above a statement about the naming, not about the design. Name the layers (or check that the repository's conventions differ from the ones this check knows) before reading a thin domain into it.
Documentation Quality: Documentation: contradicts the code
docs/brand/README.md
9.5
Near-clean
Documentation Quality: Documentation: written for insiders
How the grades work
Every finding carries one of four grades. Three say how serious it is. The fourth says this
survey could not settle it — and it is a grade, not a gap.
Critical — 15
A definite problem that already costs you something and drags the score down: a
missing authorisation check, a dependency with a known exploit, a build that does not reproduce. Failure here
tends to cause failures elsewhere.
Serious — 40
Likely wrong, but not failing yet. It degrades
the codebase over a longer horizon and can cause failures elsewhere — not urgent this week, not something to
carry for two years either.
Minor — 37
Recorded, with no effect on how the codebase functions.
Present so the survey is complete, not because it needs doing.
Could not be resolved — 1
Something this survey could not settle
from the outside, and which could be critical or serious. Either a control was required and no
positive evidence of it exists in the repository — a backup job that nothing shows was ever restored from proves
nothing about restores — or our own analysis could not run over that part of the tree. This is not a clean
result. These are excluded from the score rather than awarded a pass, so the number on the cover neither
rewards nor penalises them: if you act on this survey without resolving them, you carry that risk yourself. Each
one is named under Limitations.
Methodology & how to trust this report
Watchdog is a deep, periodic assessment — run each sprint, monthly, or quarterly, taking the time to go wider and deeper than a quick check and surfacing in one coherent report what you'd otherwise piece together from a dozen separate tools. It scores deterministically: the same commit yields the same score, every run. 75 of 83 evaluated dimensions are computed purely by tools and static analysis (confidence 1.0); 8 documentation/naming judgement(s) are LLM-assisted and labelled advisory. Overall confidence is 0.9 — the weighted average across measured dimensions; it falls as more of the score leans on LLM-assisted judgement and rises when it's fully tool-backed.
Every figure here is one of three kinds, and we label which: ✓ Measured — a deterministic fact (LoC, complexity, coverage); ~ Modeled — an estimate from a stated model (cost, effort, value-at-risk), always a range with its assumptions, never a precise fact; ◐ Advisory — an LLM prose judgement. We never present a modelled estimate as if it were measured. Perfect or absent scores carry their provenance too (ADR-0011): ✓ Tool-verified means the property itself was measured across the surface; ○ Nothing flagged means the probes came back clean — a claim bounded by what a repository can show; ⊘ Not evidenced means a working control (a tested restore, an automated rollback) showed no positive evidence — absence of evidence is not evidence of a control, so it's excluded from the score rather than awarded a spurious 10; ◐ Sampled · advisory marks an LLM verdict over a bounded sample — advisory, never a deterministic measurement.
What we checked — 83 dimensions across the health lenses
Each chip is a dimension scored from real signals across architecture, testing, dependencies, security & compliance, documentation, git-history and code quality — in one coherent pass. A surface report typically covers a handful.
How to trust any code-health report — three questions
Can you open the finding? Real findings cite a repo-relative file and line you can open at the cited line — never an absolute scratch path. Here, 58 of 92 do; the remainder are repo-wide signals — a dimension-level measurement, not a single line. (Every path in this report is repo-relative by construction: paths are normalized at the producer and the report is rejected if any rooted path leaks through.)
Is there a tool behind the number? Every score below names the method that produced it — Roslyn, git, a scanner, or (for a handful of documentation/naming dimensions) an LLM labelled sampled · advisory — not a narrative.
Does re-running give the same result? Run it again on the same commit and the score — and this report, byte for byte — is identical. A report whose numbers move between runs is describing the run, not the code.
This report answers yes to all three. That's the bar to hold any assessment to.
Tools & methods
The actual versions used this run (captured at analysis time) — re-run on the same commit for the identical score.
Method
Backs
Version
Evaluator
Roslyn static analysis
Complexity, cohesion, coupling, dead code, API surface, layering
What ran differently this time — a tool absent, degraded, or that fell back to an estimate. Named openly, not folded silently into the scores. A degraded run also records its exact cause in diagnostics.md.
D10 Test Quality — measured, with a gap in what it reached — Watchdog measured this, but not all of it. What it did not reach is a gap on our side — a collector, parser or image we have not built yet — so the numbers on that dimension cover less than the repository, and the part left out is not evidence that it would have passed. The 620 test(s) behind this row are the ones the C# collector could read, and this repository also carries at least 1 test source file(s) (.py) that it cannot: it parses C# syntax and matches C# test attributes, so a vitest/jest/JUnit/pytest-style suite is invisible to it. Skipped tests, zero-assertion tests and the other quality signals on this row are UNMEASURED in that suite — their absence from the counts above is a gap in this analyzer's language coverage, not a finding that those tests are sound.
D16 Bus Factor — not measured this run — This is a true statement about the repository that carries nothing for its owner to act on, so it is reported here rather than as a defect in their code. Single-maintainer repository — bus factor is not applicable (3 contributor(s) across 162 commit(s) sampled, automation and bot accounts excluded). One of them holds 93% of the history; the other 2 hold 3.5% each on average, below the 5% at which there is somebody to hand the work to. That is a single maintainer with drive-by contributors, not a team whose knowledge has concentrated — so the bus factor is not applicable and there is nothing here for the owner to act on.
D23 Boundary Type-Coupling — not measured this run — Watchdog could not measure this here. That is a gap on our side — a collector, parser or image we have not built yet — and it is neither a defect in this repository nor evidence that the check would have passed. At 12470 LoC across 6 projects this is a large multi-module codebase that clearly needs explicit bounded contexts. Bounded contexts cannot be inferred from a codebase that is not physically organised by them (D-321), so with none declared there are no boundaries for the coupling pass to measure across. You can widen what we reach: name this codebase's bounded contexts (≥2 module groups, e.g. per subsystem) so cross-boundary type coupling can be assessed. Declare them in `.codehealth/config.yaml` at the repository root (create it if absent), mapping each context name to the module-path or namespace prefixes that belong to it — e.g. `architecture:` → `contexts:` → `Billing: ["src/billing", "Acme.Billing"]`, `Catalog: ["src/catalog", "Acme.Catalog"]`.
D32 Data Compliance (PII/GDPR) — not measured this run — Watchdog could not measure this here. That is a gap on our side — a collector, parser or image we have not built yet — and it is neither a defect in this repository nor evidence that the check would have passed. `src/Cai.Web/ApiAccess.cs`, `src/Cai.Web/Noise/FixRateAnchor.cs`, `src/Cai.Web/Noise/MethodVersions.cs`, `src/Cai.Web/Noise/NoiseStandardEndpoints.cs`, `src/Cai.Web/Noise/NoiseStandardHealthCheck.cs`, … (+7 more) produced a parse error, so every rule in this engine's `gdpr.yml` was absent there. That absence is NOT a clean result: these rules detect personal data crossing a boundary into a log sink, a URL or browser storage, and a file that was never parsed cannot report any of the three. The rest of the tree analysed normally and its rows above stand; only these files are unaccounted for. You can widen what we reach: fix the syntax error (or exclude the file deliberately) and re-scan to cover it.
Repo exclusion declarations (.gitattributes linguist-generated/vendored, .editorconfig generated_code): none declared — every source file was scored.
Limitations & what we did not check
Watchdog assesses the repository exactly as committed, and only the repository. By design it does not reach outside the source tree: the live cloud account, the running CI/CD pipeline, the host's branch-protection and approval rules, the production configuration, or a restore actually exercised against a backup are all out of scope. That boundary is a feature, not a gap — a repo-relative, deterministic scan re-runs identically on any commit and every finding opens at a real file and line, where a live audit can neither be reproduced nor traced. The visible consequence is that controls which leave no in-repo evidence are reported as "not evidenced" and excluded from the score rather than awarded a number a static scan cannot justify.
Per-dimension blind spots
For each dimension that was measured, what a static, repo-only scan structurally cannot see — the honest edge of the measurement, not a failure of it.
D1 Cyclomatic Complexity: Cyclomatic complexity counts branches statically — it cannot tell an essential decision tree from accidental tangle, nor see complexity that lives in data or configuration (large switch-case token tables, DSL lexers/parsers, data-as-code rule tables) rather than control flow: a tokenizer's many single-character cases read as high complexity though each branch is trivial.
D2 Cognitive Complexity: Cognitive-complexity heuristics approximate how hard code is to follow; genuine domain difficulty and well-named intent that eases reading are not captured.
D3 God Classes: "God class" is sized by members and responsibilities visible in the type — a deliberately broad facade over a coherent subsystem can read the same as an accidental grab-bag. For front-end JS the file-length check is cohesion-aware (a single-responsibility module — one class/IIFE — earns a 3× threshold), but cohesion is approximated from top-level declarations, not true dependency structure.
D4 Code Duplication: Duplication is token-similarity — an in-process token-stream comparison over sliding windows, with type-aware normalization — so it finds copy-paste, not semantic duplication expressed differently. Committed machine-written code (scaffolded migrations, designer/codegen output, protobuf/OpenAPI stubs, model snapshots) is EXCLUDED — its repetition is the tool's, not the team's — so the score reflects hand-written duplication only.
D5 Coupling: Coupling is measured between projects/assemblies — runtime coupling through DI, reflection, messaging or shared databases is invisible to a static reference graph.
D6 Cohesion (LCOM4): LCOM4 cohesion is syntactic — it infers connectivity from which methods touch which fields/methods by name, not from real runtime behaviour or intent.
D7 Architectural Integrity: Layering is checked against detected/declared rules — an architecture whose boundaries live in convention or in code review, not in a rule a scanner can read, is not enforced here.
D9 Test Distribution: The test-pyramid shape is inferred from project/folder naming and references, with a single test host bucketed per-file by its path tier and content signals — a suite that names tiers unconventionally and gives no per-file signal can still be mis-bucketed.
D10 Test Quality: Assertion density is structural — it cannot tell a meaningful behavioural assertion from a trivial one, only that an assertion is present.
D11 Test Reliability: Flakiness is inferred from history/markers — Watchdog runs the suite once (for coverage), not the repeated runs under varied conditions that reveal nondeterminism, so a flaky test never recorded as failing is invisible here.
D12 Dependency Hygiene: Dependency health reads manifests and lockfiles — a vulnerability in a vendored/copied dependency, or risk from how a dependency is actually used, is outside this view.
D13 Secret Scanning: Secret detection is signature- and entropy-based on the current tree — a secret that does not match a known pattern, or one already rotated, will not be flagged (a clean scan is "nothing matched", not "no secrets exist").
D14 License Compliance: License compatibility is checked against declared package metadata and a policy — mislabelled or missing license metadata, and obligations that depend on how you distribute, are not resolved here.
D15 Churn × Complexity Hotspots: Churn hotspots come from git history — a freshly imported or squashed repository has no churn signal, and recent rewrites can mask a historically risky file.
D17 Explicit Debt: Acknowledged-debt signals (TODO/FIXME, suppressions, dead code) are textual — undocumented debt that nobody marked, and debt that lives in design rather than annotations, is invisible. Committed machine-written code (scaffolded migrations, designer/codegen output, generated stubs) is excluded — it is never the team's dead code to delete.
D18 Solution Shape: Build integrity reflects whether the solution compiled in this environment — a build that needs a private feed, a specific SDK, or a generated file absent from the repo can read as broken when it is merely unreproducible here.
D19 Documentation Quality: Documentation quality is judged by an LLM over a bounded sample of docs — it reads what is written, not whether the docs match the running system, and it is advisory, not a measurement. Its critique rows are drawn from a closed category vocabulary and each row means the same thing in every run, so two scans can be compared row by row; the SET that fires is still a sample, and does not repeat exactly. Measured on one frozen input, six scans at one engine SHA: 2-5 critique rows per scan, 8 distinct rows across the six, 3 of those 8 seen in only one scan. So a D19 row is evidence about the documentation, but a COUNT of D19 rows is not a quantity — never read a change in it as an improvement or a regression.
D20 ADR Quality: ADR quality is an LLM read of the decision records present — it cannot know about decisions made and never recorded, and its verdict is sampled and advisory.
D21 Naming Consistency: Naming quality is an LLM judgement over a bounded sample — it assesses clarity/consistency of the names it sees, not domain-correctness, and is advisory.
D22 Internal API Consistency: API-surface coherence is an LLM judgement over a sample of the public surface — consistency of intent across the whole API is approximated, not exhaustively verified.
D24 Comment Value: Comment value (WHY vs WHAT) is an LLM judgement over a bounded sample — it is advisory and cannot weigh a comment against the precise code change it was written to explain.
D25 ADR Conformance: ADR conformance is the LLM-scored fraction of sampled code that follows recorded decisions — it checks the decisions that were written down and the slices it sampled, not unrecorded rules or the whole tree.
D26 Project Cohesion: Project focus is sized from members/namespaces per project — a project that is broad by deliberate design reads the same as one that has sprawled.
D27 Navigability: Indirection/navigability is structural — it measures hops to follow a call, not whether that indirection buys real flexibility or just ceremony.
D28 Secrets (history): Secret-history scanning sweeps the git log for known patterns — a secret that predates the available history, or never matched a signature, is not found (clean means "nothing matched in the history we can see").
D29 Static Analysis (SAST): SAST findings are pattern-based (semgrep) — it finds classes of bug it has rules for; logic flaws, auth/authorization gaps and issues needing runtime context are out of reach (and clean means "no rule matched").
D30 Dependency Vulnerabilities: CVE matching depends on accurate package/version metadata and on the advisory databases — a vulnerability with no published advisory, or in code not declared as a dependency, is not seen. Coverage needs a RESOLVED graph: an unpinned requirements.txt, or a pom without a resolved build, yields partial coverage rather than a clean verdict. An ecosystem the analyzer cannot scan is reported as unmeasured, never as clean.
D34 Knowledge Freshness: Freshness is decayed commit RECENCY, not comprehension — code read often but rarely committed reads as orphaned, and stable code that genuinely needs no changes is penalised the same as forgotten code; bot/squash commits distort it like the bus factor.
D35 Change Coupling: Change coupling is co-change in COMMITS — files split across separate commits, or coupled only through a shared config/build step, read as uncoupled, and a sweeping commit (rename/format) is excluded so it doesn't couple everything. It shows that files change together, not WHY: a high coupling can be a healthy cohesive pair as readily as a hidden leak.
AC2 Forms & labels: Label association is read from static markup — a label wired up at runtime (JS-set aria-labelledby, framework-injected ids) reads as missing, a present label says nothing about whether its text is correct. A known UI-library field component (e.g. a JSX <TextField>) is now checked conservatively — flagged only when it carries NO label/aria-label/aria-labelledby/id/name — but wrapper/context-labelled libraries (Chakra/Radix FormControl+FormLabel) aren't statically visible (possible false positive) and non-JSX lowercased components are still skipped. A click handler on a plain element is now asked for a name too (it is a control the author declared), but the subtree test that answers it is deliberately generous: any DYNAMIC text expression in the subtree counts as a name, so an icon chosen by a ternary ({cond ? <IconA/> : <IconB/>}) reads as named, and a glyph component from a library the icon-import list does not know still names its parent. A clean result is "no unlabelled control found", not a labelling proof.
AC3 Page structure: Page structure is read from the static markup tree — landmarks, headings and lang injected at runtime aren't seen, heading ORDER is checked structurally (not against the rendered visual hierarchy), and lang/title/main fire only on full documents, never partials, and the data-table check sees header-cell presence (a <th> exists), not whether each header correctly associates with its cells. Static readiness, not conformance.
AC5 ARIA correctness: ARIA correctness is checked against the static role/attribute shape — roles/attributes set dynamically aren't seen, a valid role says nothing about whether it matches the element's real behaviour, and required-state checks are suppressed when a JSX spread could supply them. The two-branch toggle check (a control whose state is conveyed only by which of two mutually exclusive branches renders) reads CONDITIONALS THAT ARE ATTRIBUTES — Vue v-if/v-else/v-show and Alpine x-if/x-show — so the same toggle written as a Svelte {#if} block or a JSX ternary is control flow the markup model never projects as a branch and is not seen at all.
AC6 Visual & motion safety: Contrast and motion safety are PARTIAL by construction — literal colours (hex/rgb/hsl/named) in inline styles, in-repo <style> blocks, in-repo .css files, var() tokens, Tailwind neutral utilities and CSS-in-JS top-level declarations are read (same-rule/same-element colour+background pairs only); computed/runtime/theme colour, external-CDN stylesheets, CSS-in-JS dynamic (${…}) and nested-selector colours, cross-element pairs and image contrast stay out of reach, so a clean result is bounded by what the static CSS itself shows.
AC7 A11y enforcement: Enforcement is scored from in-repo config/CI evidence only — an a11y gate enforced in external tooling with no in-repo trace can't be credited, and a configured linter is presence, not proof the rules actually run or block a merge.
AX10 Code composition: Role is inferred from namespace/folder convention, not semantics — a domain concept living in a folder named "Services" reads as application, and the split is lines-of-code, not business value. The business-logic-share score is a SOFT, FLOORED signal: it contributes to the Architecture lens but is floored at the Critical gate, so an infrastructure-heavy design (a gateway, an ETL, a driver) is legitimately low without being nuked to zero.
M4 Documentation accuracy: Onboarding quality is an LLM read of the docs/setup present — it cannot run the onboarding or measure how long a real new joiner takes; the verdict is sampled and advisory.
P4 Deployment & Rollback: Approval/branch-protection rules live in repository settings the scan cannot see — only their in-repo evidence (config files, workflows) is checked, so a control enforced purely in the host's settings reads as "not evidenced".
P6 Release Hygiene: Rollback/observability controls are inferred from repo artefacts (pipelines, dashboards-as-code) — controls configured in external tooling, with no in-repo trace, cannot be credited.
The LLM boundary
LLM-set scores this run (8): D19, D20, D21, D22, D24, D25, D26, M4 (model: Local LLM). For these, a model reads a bounded sample and sets the numeric score; each names its own sample and method on its card. They are sampled and advisory by design: they vary at the margins between runs and are never a deterministic measurement. Every other score in this report is tool-computed at confidence 1.0.
What it measures: How tangled the control flow is — methods with many branches are hard to test and change.
Method: Cyclomatic complexity per method (1 + decision points), computed exhaustively across production source; test projects separated by convention. Deterministic.
Resolve the 1 NoiseStandardEndpoints.MapNoiseStandard (cyclomatic 133) finding(s) in Cyclomatic Complexity — start with NoiseStandardEndpoints.cs. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 NoiseSubmissions.Accept (cyclomatic 30) finding(s) in Cyclomatic Complexity — start with NoiseSubmissions.cs. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 PublicationContract.Check (cyclomatic 30) finding(s) in Cyclomatic Complexity — start with PublicationContract.cs. — One of this dimension's main actionable groups (1 warning-level).
Enforce Cyclomatic Complexity in CI to reach Verified (currently Documented). — Hardens enforcement from Documented toward Prevented — provenance only; does not change the score.
Detailed fixes: d1_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: How hard the code is for a person to follow, beyond raw branching.
Method: Cognitive complexity per method (Sonar-style nesting-penalized score), computed exhaustively over production code, excluding test projects. Deterministic.
Resolve the 1 NoiseStandardEndpoints.MapNoiseStandard (cognitive 151) finding(s) in Cognitive Complexity — start with NoiseStandardEndpoints.cs. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 PublicationContract.Check (cognitive 43) finding(s) in Cognitive Complexity — start with PublicationContract.cs. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 NoiseSubmissions.Accept (cognitive 42) finding(s) in Cognitive Complexity — start with NoiseSubmissions.cs. — One of this dimension's main actionable groups (1 warning-level).
Enforce Cognitive Complexity in CI to reach Verified (currently Documented). — Hardens enforcement from Documented toward Prevented — provenance only; does not change the score.
Detailed fixes: d2_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D3 · God Classes8.6 / 10Strong✓ Tool-verified
What it measures: Over-large classes that try to do too much ("god classes").
Method: God-class detection by line and method-count thresholds per logical type (partial classes unified), filtered for generated code and registration/contract false positives. Deterministic.
Resolve the 3 MethodTooLong finding(s) in God Classes — start with NoiseStandardEndpoints.cs, PublicationContract.cs, NoiseSubmissions.cs. — One of this dimension's main actionable groups (3 warning-level).
Resolve the 2 FileTooLong finding(s) in God Classes — start with NoiseStandardEndpoints.cs, NoiseStore.cs. — One of this dimension's main actionable groups (2 warning-level).
Resolve the 1 ClassTooLong finding(s) in God Classes — start with NoiseStandardEndpoints.cs. — One of this dimension's main actionable groups (1 warning-level).
Enforce God Classes in CI to reach Verified (currently Documented). — Hardens enforcement from Documented toward Prevented — provenance only; does not change the score.
Detailed fixes: d3_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: Copy-pasted code that should be shared instead.
Method: Code duplication via token-stream sliding windows with type-aware normalization (locals masked, type names preserved), density-scored per KLoC of production code. Deterministic.
Resolve the 1 Duplicated block (18 lines × 2) finding(s) in Code Duplication — start with NoiseStore.cs. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 Duplicated block (12 lines × 2) finding(s) in Code Duplication — start with Keys.cs. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 Duplicated block (6 lines × 2) finding(s) in Code Duplication — start with NoiseStore.cs. — One of this dimension's main actionable groups (1 warning-level).
Detailed fixes: d4_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D5 · Coupling8.7 / 10Strong✓ Tool-verified
What it measures: Whether volatile projects sit underneath others that depend on them (so their churn ripples upward), and whether project dependencies form cycles. A widely-depended-on but stable shared/kernel project is healthy, not penalised.
Method: Dependency cycles via elementary-DFS over real .csproj references, plus Martin instability (afferent/efferent) per project. Exhaustive over the reference graph, deterministic.
Coverage: Exhaustive · type-level: afferent/efferent coupling + cycles computed over every production type — the population is all types, not a name convention.
Resolve the 1 Off the main sequence finding(s) in Coupling. — One of this dimension's main actionable groups (1 warning-level).
Enforce Coupling in CI to reach Verified (currently Documented). — Hardens enforcement from Documented toward Prevented — provenance only; does not change the score.
Detailed fixes: d5_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: Whether a class's methods are focused on a single responsibility.
Method: LCOM4 cohesion per production class with at least two methods: connected components of methods sharing state or calls, computed syntactically. Deterministic, not a proxy.
Coverage: Exhaustive · type-level: LCOM4 cohesion computed over every production class — the population is all types, not a name convention.
What it measures: Whether the code respects its intended layering / architecture rules.
Method: Enforcement rung (Prevented/Verified/Documented) per checkable ADR via Roslyn, plus dependency cycles via the engine shared with D5/AX3. Deterministic, exact.
All 8 mechanizable ADR(s) are enforced: 5 by analyzers, 3 by tests. Cycles found: 0.
✓ On the Gold path — maintain.
Detailed fixes: d7_recommendation.md.
Do you agree with this assessment?
D9 · Test Distribution10.0 / 10Exemplary✓ Tool-verified
What it measures: Whether the test suite has a healthy mix of unit / integration / end-to-end tests.
Method: Test projects classified (Unit/Integration/BDD/E2E) from compiled metadata; test methods counted exhaustively across projects with placement-agnostic disk fallback. Deterministic.
620 test methods: 620 unit, 0 integration, 0 BDD, 0 e2e.
✓ On the Gold path — maintain.
Detailed fixes: d9_recommendation.md.
Do you agree with this assessment?
D10 · Test Quality10.0 / 10Stronggated by 2 serious findings✓ Tool-verified
What it measures: Whether the tests truly assert behaviour rather than just running the code.
Method: Per-test assertions, skips, and mock references analyzed via Roslyn; structured skip-reason tags (BUG:/ENV:) separate documented deferrals from debt. Deterministic.
0 skipped, 0 zero-assertion, 2 depending on a live external host, no mocking-framework packages referenced (hand-written doubles or no mocking) across 620 tests. Measured on the C# suite only — at least 1 test source file(s) (.py) went unread, so its test quality is unmeasured and is not in these counts.
Depends on a live external host: A_marketing_origin_may_call_the_api_from_a_browser · ×2tests/Cai.Tests/PublicCorsTests.cs:28
What to do
Resolve the 2 Depends on a live external host finding(s) in Test Quality — start with PublicCorsTests.cs (2). — One of this dimension's main actionable groups (2 warning-level).
Detailed fixes: d10_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D11 · Test Reliability10.0 / 10Exemplary✓ Tool-verified
What it measures: Whether the tests pass reliably, with no flakiness.
Method: Suite re-run N times within tiered wall-clock budgets (unit to e2e); tests failing non-deterministically across runs flagged; guarded tests retried when #if guards detected.
What it measures: Whether dependencies are current, secure, and not bloated.
Method: Manifest scan via dotnet list package across all projects; worst-signal-per-package deduction (saturating for vulnerabilities, capped-linear for deprecation/outdated) per KLoC. Exhaustive, deterministic.
What it measures: Whether any secrets (keys, tokens, passwords) have leaked into the code.
Method: In-process native secret scanner (entropy plus signature patterns) across all tracked files; no external tool. A clean result is a measured 10, not no-data zero. Deterministic.
Resolve the 1 Leaked secret finding(s) in Secret Scanning — start with cai-delivery.sample-key.json. — One of this dimension's main actionable groups (1 issue-level).
Enforce Secret Scanning in CI to reach Verified (currently Documented). — Hardens enforcement from Documented toward Prevented — provenance only; does not change the score.
Detailed fixes: d13_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: Whether the licenses of third-party packages are compatible with your policy.
Method: Third-party package licenses resolved from declared package metadata and checked against the configured policy (allow/deny/copyleft). Deterministic; clean = no incompatible license found at metadata depth.
What it measures: Files that change often and are also complex — the riskiest hotspots.
Method: Per production file churn times cyclomatic complexity over a rolling window, computed from git and Roslyn/JS/Razor analysis. Exhaustive, deterministic per commit date.
Resolve the 5 Hotspot finding(s) in Churn × Complexity Hotspots — start with NoiseStandardEndpoints.cs, NoiseSubmissions.cs, CaiScorer.cs. — One of this dimension's main actionable groups (5 warning-level).
Resolve the 1 Repeated repair finding(s) in Churn × Complexity Hotspots — start with ApiRateLimiting.cs. — One of this dimension's main actionable groups (1 warning-level).
Detailed fixes: d15_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: Acknowledged debt left in the code — TODOs, dead code, suppressed warnings.
Method: Roslyn syntactic debt markers (suppressions/TODO/FIXME/HACK/empty-catch/commented-code/Obsolete) plus SymbolFinder dead-code analysis; weighted-debt-per-KLoC density deducted 2.0x per unit. Deterministic, exhaustive.
What it measures: Whether the solution is laid out in a sensible, conventional structure.
Method: Solution structure: project count, decomposition, shell-project detection, build success (confirmed failures cap the score); traced to actual .sln files and binaries. Deterministic.
6 projects, 127 source files, 24402 hand-written lines of code (12470 production / 11932 test — the split is derived per file: test is what its project, its own path, or a compile-guarded region marks as test, and a file carrying no test signal counts as production), 9 inter-project edges.
What it measures: Whether the project's documentation is clear, complete, and useful.
Method: Judged by language model at low temperature (0.0-0.1) on a deterministic doc sample (READMEs plus first 25 architecture docs), with two-pass stability filtering. Advisory, sampled.
This is a comprehensive CAI project with an open, reproducible standard documented across READMEs and architecture docs: the CAI itself (with bands, repository layout, the reference scorer), its web UI (Blazor static-SSR + API endpoints), the scoring library (its deterministic fold, lens weights, headline reconstruction), the CLI proof-of-concept, unattested rubric catalogs explaining why withheld, a brand mark and its asset references, architecture decisions as ADRs, benchmarks for the hot paths, and an authoritative challenge-and-review document. The XML-doc coverage is strong: Cai.Benchmarks at 90%, Cai.Scoring at 100%, and all four CAI docs are covered. This repository's documentation is exemplary: a well-organized set of architecture decision records (ADRs) under docs/adr/, each one a self-contained numbered entry describing context, the architectural choice, and its consequences; plus a dedicated format-specification document for CAI-delivery packages. The ADRs cover the project's design decisions (signed delivery package, deterministic scoring, free/paid firewall), while the format doc explains signed evidence bundles, verdicts, signatures, signing canonicalization, offline verification, keys/rotation, versioning, and security considerations. There is no missing-installation or usage content among the documents shown; all are well-structured with headings and an outline present in the visible text.
Documentation: contradicts the code · ×3docs/CHALLENGE.md
Detailed fixes: d19_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D20 · ADR QualityExemplary◐ Sampled · advisory
What it measures: Whether architecture decisions are recorded well (context, decision, consequences).
Method: Per-ADR judgment by language model at low temperature with two-pass stability; confidence is share of ADRs evaluated; enforcement-field presence detected deterministically. Advisory.
What it measures: Whether names — types, methods, variables — are clear and consistent.
Method: Judged by language model at low temperature (0.0-0.1) on a deterministic random symbol sample (fixed size, not exhaustive), with disclosed confidence band. Advisory, sampled.
1 naming inconsistencies across 200 sampled symbols.
The property `Ct` is used in two different test classes. While `Ct` is a common abbreviation for 'Context' or 'Count' in testing frameworks, using the same short, ambiguous abbreviation for potentially different concepts (e.g., a test context object vs. a count of items) reduces readability. It is preferable to use descriptive names like `TestContext` or `ItemCount`.
✓ On the Gold path — maintain.
Detailed fixes: d21_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D22 · Internal API ConsistencyStrong◐ Sampled · advisory
What it measures: Whether the internal API surface is consistent and coherent.
Method: Judged by language model at low temperature over a sample of the public API surface (IsPackable or .Contracts types). Sampled, advisory; confidence discounted by model uncertainty.
Duplicate intent: Both `Bands` and `BandCutlines` provide a `For` method that maps a score to a `Band`. `BandCutlines` is a property of `ScoringParameters` and likely contains the specific cutlines used for a version, while `Bands` appears to be a static utility. This creates ambiguity on which class to use for score-to-band conversion.
Confusingly similar names: `Sign` and `SignPackage` are distinct operations (one returns `DeliverySignature`, the other `DeliveryPackage`), but the naming convention suggests `SignPackage` might be a variant of `Sign` rather than a distinct higher-level operation. It is unclear if `Sign` is a low-level primitive or if `SignPackage` is the primary entry point.
Ambiguous boolean semantics: `SignatureValid` checks cryptographic integrity, while `AuthenticAndReproducing` likely checks semantic correctness (score reproduction). However, `Reproduced` is also a property. The naming `AuthenticAndReproducing` is a compound boolean that obscures whether it is a logical AND of `SignatureValid` and `Reproduced` or a separate check. If it is derived, it should be a method or clearly documented as a composite property.
What to do
Resolve the 1 Duplicate intent finding(s) in Internal API Consistency. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 Confusingly similar names finding(s) in Internal API Consistency. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 Ambiguous boolean semantics finding(s) in Internal API Consistency. — One of this dimension's main actionable groups (1 warning-level).
Detailed fixes: d22_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D24 · Comment ValueExemplary◐ Sampled · advisory
What it measures: Whether comments are worth it — explaining WHY (valuable) rather than WHAT (redundant).
Method: Judged by language model at low temperature (0.0-0.1) on deterministically sampled inline comments with surrounding code; findings verified back to sampled comments by substring match. Advisory, sampled.
What it measures: Whether the code actually follows the decisions recorded in the project's ADRs.
Method: Judged by language model at low temperature against ADRs plus a deterministic structural code summary; findings linked to repo-rooted ADR paths for traceability. Advisory.
1 of 6 projects flagged as possibly oversized/incoherent.
Projects may be oversized for their cohesion
What to do
Resolve the 1 Projects may be oversized for their cohesion finding(s) in Project Cohesion. — One of this dimension's main actionable groups (1 recommendation-level).
Detailed fixes: d26_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: How far you must trace to follow a call — low indirection and co-located slices read easier.
Method: Call indirection (interface hops, cross-namespace calls, slice-locality scaled) over a sampled set of method invocations, size-aware baseline. Sampled; confidence discounted by symbol-resolution gaps.
Coverage: Slice locality from the first namespace segments, SAMPLED (≤400 methods) — not exhaustive.
79 % of calls cross a namespace and 2 % go through an interface, but 100 % of collaborators are co-located — so a call's collaborators sit together and tracing stays easy. Baseline: medium — clean/modular boundaries expected.
What it measures: Whether any secrets were ever committed — scanned across the full git history, not just now.
Method: Secret scan via TWO gitleaks detect passes in an isolated checkout — the full git history, then a second --no-git pass over the working tree as it stands — merged and de-duplicated by (rule, file, line); each match flagged High. Both invocations are recorded in the audit trail. Exhaustive; when the tool is absent, or when its output cannot be parsed into the expected shape, the dimension is WITHHELD as an explicit measurement gap on our side — unscored and excluded from the lens, never a hedged middling score.
5 finding(s): 0 critical, 5 high, 0 medium, 0 low. Remediation for historically-committed secrets is credential rotation — they remain in history regardless of later deletion.
Rotate the exposed credentials — git history can't be un-committed
What to do
Resolve the 5 Secret finding(s) in Secrets (history) — start with cai-delivery.keys.json (3), cai-delivery.sample-key.json (2). — One of this dimension's main actionable groups (5 issue-level).
Resolve the 1 Rotate the exposed credentials finding(s) in Secrets (history). — One of this dimension's main actionable groups (1 recommendation-level).
Detailed fixes: d28_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: Real static-analysis (SAST) findings — likely security bugs in the code, any language.
Method: Polyglot static analysis via semgrep across the repo using the pinned, image-baked p/security-audit + p/owasp-top-ten rulesets (no scan-time registry fetch); severity rules (ERROR/WARNING/INFO) map to a full-band severity-weighted score. Exhaustive, deterministic; degrades on parse failure.
Coverage: semgrep pattern rules over all files — exhaustive for the rule set, blind to classes of bug without a rule (clean = no rule matched).
16 finding(s): 0 critical, 9 high, 3 medium, 4 low. semgrep hit a parse error in 13 file(s) — `src/Cai.Web/ApiAccess.cs` (line 16), `src/Cai.Web/Noise/FixRateAnchor.cs` (line 94), `src/Cai.Web/Noise/MethodVersions.cs` (line 56), `src/Cai.Web/Noise/NoiseStandardEndpoints.cs` (line 803, line 2693), `src/Cai.Web/Noise/NoiseStandardHealthCheck.cs` (line 25), … (+8 more) — so no absence of findings in the named regions is evidence of anything; rows reported elsewhere in those files are real. Fix the syntax error (or exclude the file deliberately) and re-scan to cover them. In 5 of those file(s) — `src/Cai.Web/ApiAccess.cs`, `src/Cai.Web/Noise/NoiseStandardHealthCheck.cs`, `src/Cai.Web/Registry/RegistryAuthentication.cs`, `src/Cai.Web/Registry/RegistryHealthCheck.cs`, `src/Cai.Web/RubricsHealthCheck.cs` — the break is a C# primary constructor, which semgrep's grammar cannot parse and which hides the type from every rule scoped to it; those files were re-scanned from a shadow copy with the constructor's parameter list blanked (line and column preserving), which restores the type to the rules, and 0 additional row(s) came from that pass. Only the header's own parameter list is unread in them. In 2 of those file(s) — `src/Cai.Web/ApiAccess.cs`, `src/Cai.Web/Noise/NoiseSubmissions.cs` — the break is at a TYPE DECLARATION (a C# primary constructor semgrep's grammar cannot parse), so the loss is wider than the named lines: rules scoped to that type see no type to scope to and are blind over its whole body, while rules matching statements keep working there. Absence of a type-scoped finding in those types is not evidence of anything. Separately, one or more rules could not re-parse an embedded snippet in 2 file(s) (e.g. a workflow `run:` block read as shell). Those files WERE scanned and their other rows are unaffected; only those rules' view of those snippets is missing.
High: dependabot-missing-cooldown · ×9.github/dependabot.yml:7detected by semgrep finding
Medium: watchdog-eol-node-runtime-in-workflow · ×3.github/workflows/a11y.yml:29detected by semgrep finding
Low: request-host-used · ×4deploy/nginx/api.codeassuranceindex.info.conf:16detected by semgrep finding
What to do
Resolve the 9 High finding(s) in Static Analysis (SAST) — start with publish-packages.yml (4), NoiseStore.cs (2), RegistryStore.cs (2). — One of this dimension's main actionable groups (9 issue-level).
Resolve the 4 Low finding(s) in Static Analysis (SAST) — start with codeassuranceindex.info.conf (2), api.codeassuranceindex.info.conf, app.codeassuranceindex.info.conf. — One of this dimension's main actionable groups (4 recommendation-level).
Resolve the 3 Medium finding(s) in Static Analysis (SAST) — start with RubricCatalogStore.cs (2), a11y.yml. — One of this dimension's main actionable groups (3 warning-level).
Detailed fixes: d29_recommendation.md · top locations in Appendix A, every location in findings.md.
What it measures: Whether any dependency has a known published vulnerability (CVE), direct or transitive, in ANY ecosystem the repository declares — Dart pub, Elixir/Hex, Go modules, Java and Kotlin via Maven/Gradle, JavaScript/npm, .NET/NuGet, PHP/Composer, Python/PyPI, RubyGems, Rust/Cargo and Swift.
Method: Dependency-CVE scan across every ecosystem the repository declares, scored ONCE. Three sources are unioned and deduplicated by advisory identity (rule id + alias closure, CVE<->GHSA) scoped to package+version, keeping the worst severity: `osv-scanner --recursive` over osv.dev for Dart pub, Elixir/Hex, Go, Java and Kotlin via Maven/Gradle, npm, PHP/Composer, Python/PyPI, RubyGems, Rust/Cargo and Swift; `trivy fs --scanners vuln` for npm lockfiles; and `dotnet list package --vulnerable --include-transitive` for NuGet (with per-advisory collapse of the project x target-framework fan-out), plus a DECLARED-dependency arm that resolves a published gem's gemspec against rubygems.org where no Gemfile.lock is committed. `SeverityScore(c,h,m,l, normalizer 8.0)`. NotApplicable only when NO ecosystem is readable; if any applicable ecosystem could not be scanned the findings are REPORTED and the score is withheld. Supersedes the npm and OSV arms, retired 2026-09-05.
What it measures: Whether anyone still has living knowledge of each file, or it has been orphaned — last understood long ago by someone now gone quiet. The sibling of the bus factor: D16 asks who owns it, D34 asks whether anyone still knows it.
Method: File orphaning as total living-knowledge decay below one focused-commit's worth within a year, computed per-file from the D16 decay model. Exhaustive, deterministic over fixed history.
Every significant source file has living knowledge — recently and meaningfully worked. Counted over 57 of the 66 production source files in this repository: the rest are under the ~2,400-byte size floor this dimension measures over.
What it measures: Whether files that change together actually belong together — pairs that repeatedly co-change in git history despite having no explicit code dependency, surfacing the hidden/logical coupling (and boundaries in the wrong place) a static scan can't see.
Method: Pairwise co-occurrence over the per-commit file sets in git history (production source only — tests and generated dropped): Degree-of-Coupling = shared ÷ min individual revisions, reported above noise floors (each file ≥10 revisions, ≥5 shared commits, ≥50% strength); sweeping commits excluded. Deterministic over fixed history.
Coverage: Population: PRODUCTION source files only — test and generated files are dropped before pairing, so a class co-changing with its own test (trivially ~100%) can't drown the real production↔production coupling. Pairs ranked by Degree-of-Coupling. A non-source file is never a coupling PARTICIPANT either: documentation, schemas, config and data files are dropped with the rest, so a code↔docs pair — a command and the reference page that restates it — is not reported however strongly the two co-change; nor is coupling that runs THROUGH a build step or config file.
What it measures: Whether the build pipeline provides supply-chain integrity — generated provenance/attestation, signed artifacts (cosign/sigstore), an SBOM, and pinned build actions. Presence of the configuration, not a runtime guarantee.
Method: Supply-chain provenance/signing read deterministically from CI/build config (.github/workflows, .gitlab-ci.yml, azure-pipelines, Jenkinsfile, .circleci) + the release surface: four signals — generated provenance/attestation (SLSA/in-toto/actions-attest), artifact signing (cosign/sigstore/gitsign), an SBOM (syft/sbom-action/*.spdx.json/*.cdx.json), and SHA-pinned build actions — scored 10·present/denom. NotApplicable without a build pipeline. Detects configuration presence, not runtime enforcement.
Build pipeline generates provenance, signs artifacts and produces an SBOM.
Secret passed as a command-line argument
Release publish has no approval gate
What to do
Resolve the 1 Secret passed as a command-line argument finding(s) in Supply-chain Provenance & Signing. — One of this dimension's main actionable groups (1 warning-level).
Resolve the 1 Release publish has no approval gate finding(s) in Supply-chain Provenance & Signing. — One of this dimension's main actionable groups (1 recommendation-level).
Detailed fixes: d36_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
D39 · IL Efficiency9.6 / 10Exemplary✓ Tool-verified
Method: IL instruction count per method, read from the BUILT first-party assemblies via Mono.Cecil (the target is compiled on a deep run); scored on the fraction of methods whose emitted IL body exceeds the size threshold. Sees compiler-generated bloat source can't; not-applicable when the target fails to build. Deterministic.
8 of 443 first-party methods have an oversized IL body.
IL efficiency: 8 authored method(s) exceed the IL budgetsrc/Cai.Web/Noise/NoiseSubmissions.cs:153
✓ On the Gold path — maintain.
Detailed fixes: d39_recommendation.md · top locations in Appendix A, every location in findings.md.
Do you agree with this assessment?
Frontend & cross-cutting dimensions
R = React/JS · M = Maturity · P = Readiness.
AC2 · Forms & labels6.8 / 10Strong✓ Tool-verified
Other · Accessibility — Whether form controls have a programmatic label (an associated label, aria-label or aria-labelledby), buttons have text, links have an accessible name, a click handler on a plain element names the control it declares, fieldsets have a non-empty legend, known UI-library field components carry a label prop, and a placeholder isn't used as the only label. Static markup readiness, not a WCAG conformance claim.
Method: Static markup-model scan: inputs/selects/textareas checked for an associated label[for]/wrapping label/aria-label/aria-labelledby (per document), buttons for accessible text, fieldsets for a legend; placeholder-only labelling flagged. Deterministic, hard fact per control.
Coverage: Population: form controls, buttons, links, fieldsets and known UI-library field components in the PARSED MARKUP files only (.html/.htm/.cshtml/.razor/.vue/.svelte/.jsx/.tsx); components, hidden subtrees and spread/dynamic-attribute elements are skipped, so a control whose label arrives through a spread or a runtime expression is deliberately not judged. Markup built in script — tagged-template (html`…`) UIs and hyperscript DOM factories — is not read at all.
A fieldset groups related controls but has no <legend> to name the group. Add a <legend> as its first child — or, if the group already has a visible caption beside it (or the fieldset is there only to disable its subtree and carries no group box), point aria-labelledby at that caption's id instead, which names the group without rendering a second one. — src/Cai.Web/Components/Pages/NoiseRate.razor:183
What to do
Give every control a programmatic label (a <label for> / wrapping <label> / aria-label) and every button text — a placeholder is not a label.
Other · Accessibility — Whether pages declare a language (well-formed BCP-47) and a non-empty title, expose exactly one main landmark and a sane heading order with non-empty headings, keep zoom enabled, title their iframes, give data tables header cells, and avoid meta-refresh. Static markup readiness, not a WCAG conformance claim.
Method: Static markup-model scan: html lang, document <title>, a main landmark and heading order on full documents only, plus zoom-disabling viewports, untitled iframes and meta-refresh anywhere. Deterministic, per structural checkpoint.
Coverage: Population: the PARSED MARKUP documents (.html/.htm/.cshtml/.razor/.vue/.svelte/.jsx/.tsx). The page-level checks — lang, title, single main landmark — fire ONCE PER FULL DOCUMENT (an <html> root) and never on a partial or component fragment, so a repo of fragments is assessed only on the per-element checks (heading order, table headers, iframe titles, meta-refresh, zoom). Markup built in script — tagged-template (html`…`) UIs and hyperscript DOM factories — is not read at all.
Other · Accessibility — Whether ARIA is used correctly — valid non-abstract roles, the ARIA state a role requires, valid (non-misspelled) aria-* attribute names, in-enum values for token-typed aria-* attributes, and no aria-hidden on (or wrapping) a focusable element. Static markup readiness, not a WCAG conformance claim.
Method: Static markup-model scan: role values checked against the WAI-ARIA role set (abstract/invalid flagged), required ARIA state for a role, and aria-hidden on a focusable element. Deterministic, role/attribute level.
Coverage: Population: elements in the PARSED MARKUP files (.html/.htm/.cshtml/.razor/.vue/.svelte/.jsx/.tsx) that carry a role or an aria-* attribute; roles and token values are checked against the ARIA enums exhaustively within that set. An expression-valued (dynamic) role or aria-* value is skipped rather than guessed, and markup built in script — tagged-template (html`…`) UIs and hyperscript DOM factories — is not read at all.
Other · Accessibility — Whether focus outlines aren't removed without a replacement, motion respects prefers-reduced-motion, and literal CSS colour pairs meet contrast — PARTIAL: inline styles, in-repo <style> blocks, in-repo .css files, var() tokens, Tailwind neutral utilities and CSS-in-JS literals are read (hex/rgb/hsl/named), never computed/runtime/external-CDN colour. Static markup readiness, not a WCAG conformance claim.
Method: Static markup/CSS scan: inline outline:none/0, literal inline colour/background contrast against the 4.5:1 AA floor, and <style>-block animation without a prefers-reduced-motion guard. Deterministic but PARTIAL — only inline styles and in-repo CSS literals are visible.
Coverage: Population: styled elements in the PARSED MARKUP files (.html/.htm/.cshtml/.razor/.vue/.svelte/.jsx/.tsx), plus in-repo <style> blocks, in-repo .css files and CSS-in-JS literals. Colour contrast is computed from LITERAL colour pairs only (hex/rgb/hsl/named, including var() tokens and Tailwind neutral utilities) — computed, runtime-themed and external-CDN colour is never resolved, so this is a partial read of contrast by construction. Markup built in script — tagged-template (html`…`) UIs and hyperscript DOM factories — is not read at all.
Other · Accessibility — Whether accessibility is ENFORCED in the toolchain — an accessibility checker configured over the markup (an a11y lint rule set, e.g. eslint-plugin-jsx-a11y or vuejs-accessibility where the project lints JavaScript) and an automated accessibility assertion wired into tests or CI (axe/pa11y/Lighthouse or an equivalent) — on the Documented→Verified→Prevented ladder.
Method: Repo config/CI scan: an accessibility checker configured over the markup (an a11y lint rule set such as eslint-plugin-jsx-a11y / vuejs-accessibility where JavaScript is linted) and an automated accessibility assertion in tests or CI (axe/pa11y/Lighthouse or equivalent), graded on the Documented→Verified→Prevented rungs. Deterministic, presence/rung detection.
Coverage: Population: the repository's own tooling configuration — lint config, test and CI files — NOT the markup. It is read for a configured accessibility checker and an automated accessibility assertion (axe/pa11y/Lighthouse, or a native-toolkit equivalent), and it credits an INVOCATION, never a mention: a licence filename, an import comment or a doc reference earns no rung. Enforcement configured entirely outside the repository leaves no evidence here and cannot be credited.
Other · Architecture — Whether any singleton service captures a scoped/transient dependency — a silent lifetime/threading bug.
Method: Roslyn scan: DI registrations parsed from AddSingleton/Scoped/Transient; each singleton checked for captured shorter-lifetime dependencies. Exhaustive, deterministic.
Other · Architecture — How the codebase splits by code ROLE — domain, application, infrastructure, test, generated. The significance map behind the knowledge/coupling weighting, and a DDD signal in its own right: a thin domain core under fat infrastructure is the anemic-domain smell, quantified. How each file's role is decided, because the split is only as good as that: a generated name or a build-output tree makes it Generated, a test project makes it Test, and otherwise the file's NAMESPACE and PATH words are matched against fixed vocabularies in a fixed ORDER — domain, then infrastructure, then application — so a file whose words hit two layers is counted under the earlier one. A production file matching none of them counts as application, so that share reads 'application or unclassified' rather than a measured application layer. Roles come from naming convention, never from what the code does.
Method: Roslyn line-count by code ROLE: every source file classified Domain/Application/Infrastructure/Test/Generated by namespace + path convention (the shared CodeRoleClassifier), then significant lines summed per role. Deterministic; the advisory score is the business-logic (domain+application) share of production code.
Coverage: Population: ALL source files, each bucketed into ONE of five roles (Domain/Application/Infrastructure/Test/Generated) by namespace + path convention — a file whose layer isn't named in the convention falls to Application (the neutral default), and the split is line-count, not semantic depth or business value.
What to do
The domain core is a small share of production code, but most of the rest matched no layer vocabulary at all — so this is not yet an anemic-domain finding. The namespace/path convention could not place that code, which makes the composition above a statement about the naming, not about the design. Name the layers (or check that the repository's conventions differ from the ones this check knows) before reading a thin domain into it.
Other · Architecture — Whether singleton services avoid mutable shared instance state that concurrent callers would race on.
Method: Roslyn scan: singleton field mutations unguarded by lock or Interlocked, per type; syntax-based guard detection. Deterministic, traceable per field.
Other · Architecture — Whether the project-reference graph is acyclic (cycles block independent build/deploy and signal eroding boundaries).
Method: Project reference cycles via elementary-DFS over real .csproj references, using the engine shared with D5/D7; cyclic versus acyclic. Exhaustive, deterministic.
Other · Architecture — Whether dependencies point inward (Domain ← Application ← Infrastructure/Web) — the clean-architecture dependency rule, checked across the project graph.
Method: Layer violations by name-segment inference (Domain/Core to Application to Infrastructure/Web) over the project-reference graph. Exhaustive over all projects, deterministic.
Other · Architecture — Whether the codebase has a recognisable, scale-appropriate structure (a named architectural style, or modular enough for its size) rather than being an ad-hoc ball of mud.
Method: Roslyn plus csproj analysis: architecture style detection (DDD, clean, vertical-slice, CQRS) and structure fitness for repo size. Deterministic.
Other · Architecture — Whether interfaces stay focused rather than fat — the Interface-Segregation principle (SOLID 'I').
Method: Roslyn scan: public interface declared-member counts (accessors fold into their property/event); fat-interface threshold (over 15 declared members) flagged per type. Each finding also reports the distinct-OPERATION count — members counted by name, so an overload group counts once — which decides whether it states the caller-side ISP harm or the implementer-side burden of an overload set. Deterministic, type-level.
`INoiseStore` declares 28 members: `TryRecordSubmission`, `FindSubmission`, `AlreadySubmitted`, `ListSubmissions`, `ConfigurationJson`, `RecordVerdict`, `RecordResolution`, `RegisterPrompt`, `ListVerdicts`, `ListResolutions`, `ListPrompts`, `RecordPublication`, `LatestPublication`, `PublishedPeriods`, `PublishedTallies`, `RecordRejudge`, `ListRejudge`, `RaiseDispute`, `FindDispute`, `ResolveDispute`, `ListDisputes`, `RegisterIntent`, `ListIntent`, `RecordFindings`, `FindFinding`, `RecordCost`, `CostFor`, `JudgedPeriods`. Counted as the author wrote them — a property is ONE member and its get/set accessors are not counted separately, and an event counts once. A wide interface forces every implementer and caller to depend on methods they don't use (the Interface-Segregation 'I' in SOLID). Split it into focused role-interfaces. — src/Cai.Web/Noise/NoiseStore.cs:79
What to do
Split fat interfaces into focused role-interfaces so clients depend only on what they use.
Do you agree with this assessment?
AX8 · Test isolation10.0 / 10Exemplary✓ Tool-verified
Other · Architecture — Whether production projects stay free of references to test projects — tests may depend on production, never the reverse.
Method: Csproj graph: each production project checked for references to test projects (identified by test-framework presence, not name). Zero violations is clean. Deterministic.
Do you agree with this assessment?
C2 · Access Controls8.0 / 10Strong✓ Tool-verified
Other · Security — Whether access is authorized by default — a framework authorization attribute/decorator or policy, or imperative guard methods (throw-on-violation) called from handlers.
Method: Roslyn scan: [Authorize] usage and authorization policies, plus imperative throw-on-violation guard methods detected via syntax. Deterministic.
[AllowAnonymous] (15) outweighs [Authorize] (1) — review whether the open surface is intended.
What to do
Review the [AllowAnonymous] surface — keep anonymous endpoints deliberate and documented so the open surface stays intentional.
Other · Code Health — Unreviewed-generation residue: shipped members still throwing NotImplementedException, and placeholder string literals left in non-test, non-generated code. Scored as a quality signature, never as a claim about authorship.
Method: Roslyn syntax scan: NotImplementedException throws and placeholder string literals in non-test, non-generated shipped code. Deterministic, code-shape signature.
Other · Code Health — Unfinished work detected by code SHAPE, not keywords: members that only throw a "not implemented" exception, methods that take inputs and return a constant, async methods that never await, dead `if (false)` / `#if false` branches, and skeleton types most of whose members are holes. A real, objective slice of technical debt.
Method: Roslyn syntax scan: incompleteness by code shape (constant-returning methods, async-never-await, #if false branches, guards that return what the code already falls through to, tests an earlier guard already decided, comparisons against NaN, skeleton types), not keyword-gated. Deterministic, code-shape heuristic.
Maturity · Maturity — Whether the repo and its projects have a README, and whether it's substantive and current.
Method: Filesystem scan: README presence, word count, and headings for depth; git history for staleness. Exhaustive across root and project dirs, deterministic.
What to do
Add a build/run (quick start) section to the root README — the first thing a newcomer needs.
Add a 'Testing' section to the root README — how to run the test suite.
Add a README to the 1 of 6 project(s) that lack one — worth up to 0.3 pts.
Maturity · Maturity — Whether the README actually describes the code that exists (LLM-judged, advisory).
Method: Judged by language model at low temperature: README accuracy versus actual projects, within a disclosed tolerance. Advisory, not a measured number.
Readiness · Readiness — Whether an automated pipeline builds and tests every change.
Method: Filesystem scan: CI workflow files (.github/workflows, .gitlab-ci.yml, etc.) for build and test stages. Exhaustive, deterministic.
Do you agree with this assessment?
P10 · Library API & versioning6.0 / 10Adequate✓ Tool-verified
Readiness · Readiness — For a library: a deliberate (small) public API surface and explicit semantic versioning so consumers can depend on it safely.
Method: Roslyn scan: public API surface area and semantic-versioning markers (SemVer attributes, changelog entries) for libraries. Exhaustive, deterministic.
187/207 types (90%) are public. For a library, every public type is a stability contract — make internal-by-default and expose only the intended API.
What to do
Make types internal by default; expose only the deliberate public API so internals can change without breaking consumers.
Only 2/3 service-like projects use logging (pure contract/DTO projects are excluded — they have nothing to log). Of those 3, 2 ship a process this repository operates; the rest are libraries their consumer hosts, where the logging decision belongs to the host.
What to do
Extend structured logging across the projects you operate, and give the library ones a diagnostics seam instead — an `EventSource`/`ActivitySource` the host can subscribe to, or an optional logger on your options object — rather than taking a logging dependency on your consumers' behalf.
Readiness · Readiness — Whether SAST, secret/dependency scanning and performance benchmarking are wired in (presence, not runtime).
Method: Filesystem scan: SAST configuration, dependency-update automation, secret scanning, and a benchmark harness or benchmark step — in this repository's own ecosystem. Exhaustive, deterministic.
What to do
Add gitleaks/trufflehog in CI to block PRs that introduce committed secrets.
Readiness · Readiness — Whether releases are automated and safely reversible (probes, rolling updates, approval gates) — from manifests/pipeline files, not the live environment.
Method: Filesystem scan: deployment manifests/IaC (K8s YAML, Helm, Terraform) for rolling updates, probes, approval gates, migration hooks. Exhaustive, deterministic.
Readiness · Performance — Whether the code protects its performance with benchmarks — a benchmark suite, allocation/memory measurement, and (ideally) a CI gate. Presence is credited as a bonus, never a deduction.
Method: Repo + source scan: BenchmarkDotNet referenced (csproj/source), [Benchmark]/[MemoryDiagnoser] attribute counts, and a benchmark step in CI — scored as a bonus ladder (absence is neutral, never a deduction). Deterministic, presence detection.
Readiness · Performance — Whether the code is written to minimise allocations so it doesn't pressure its host's memory manager — buffer/slice views over copies, object pooling, stack or value-type allocation, and buffer writers. Reward-only: credited where present, never penalised where a simpler style is fine.
Raise allocation-aware density on the hot paths — currently 35 use(s) across 12,601 production line(s) (~2.8/1k). More Span/Memory, pooling (ArrayPool/ObjectPool), stackalloc and ValueTask on the allocation-heavy paths climbs this toward 10.
Readiness · Performance — Whether asynchronous code stays responsive — it avoids sync-over-async blocking (.Wait()/.GetAwaiter().GetResult()) that wastes threads and risks deadlock, and, where the code is a reusable library, awaits with ConfigureAwait(false) so it never captures and stalls its caller's context.
Method: Production-source scan: sync-over-async blocking (.Wait()/.GetAwaiter().GetResult()) counted everywhere, and — for a library with ≥5 awaits — the share of awaits using ConfigureAwait(false). Deterministic, syntax/text detection.
Other · Security — Transport security, security headers, secure cookies, input validation, middleware order and crypto hygiene (presence, not runtime).
Other · Code Health — Whether the code avoids sync-over-async (deadlock-prone blocking on tasks) and async void.
Method: Roslyn syntax scan: async methods scanned for .Wait()/.GetAwaiter().GetResult() and async-void outside event handlers. Deterministic, hard fact per invocation.
Other · Code Health — Whether any branch is dead by construction — a switch arm whose label can never equal a case-normalised subject, or an `else if` whose predicate the arm above has already swallowed.
Method: Roslyn syntax + semantics: switch labels compared against the subject's own case normaliser, and if/else-if chains checked for a literal an earlier arm's containment test already swallows. Deterministic, provable per finding. Advisory.
Do you agree with this assessment?
X13 · Undrained process stream10.0 / 10Exemplary○ Nothing flagged
Other · Code Health — Whether a child process that has BOTH standard streams redirected drains both — reading one to the end while the other is never read deadlocks once the child fills the unread pipe.
Method: Roslyn syntax + semantics: ProcessStartInfo launches with both streams redirected, checked for a drain of each stream across the enclosing type. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a loop that shortens a string until it fits a length budget has a floor — one with none grinds the value down to the empty string, or past it into a negative-length `Substring`.
Method: Roslyn syntax + semantics: while/do loops whose body's only effect on a string is to drop its last character, checked for whether anything — a direct comparison on the length, a body guard, a break — bounds that length below. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a type's disposal matches what it OWNS — releasing what it created, leaving alone what it was handed, and not declaring a finalizer for state that has nothing unmanaged to finalize.
Method: Roslyn syntax + semantics: every assignment to a disposable field is read to decide whether the type CREATED the value or was handed it, and the type's disposal is checked against that answer — an injected interface it disposes, a value it constructed and never releases, a finalizer on a type holding nothing unmanaged, and a disposable local whose every reference is a plain member read. A value handed to a container that disposes its contents (a parent control's `Controls` collection, a component `IContainer`) is released by that container and is not reported; generated code is out of population. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a method that temporarily changes state belonging to the whole process — the working directory, an environment variable — puts it back on EVERY path: a restore reached only when nothing throws leaks the change to the rest of the process.
Method: Roslyn syntax + semantics: method bodies that write the process working directory or an environment variable and write it back in the same body, checked for whether that restore sits in a `finally`/`catch` or only on the straight-line path. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether async methods accept a CancellationToken so work can be cancelled (adoption curve).
Method: Roslyn scan: every async method (excluding framework-fixed overrides/Blazor handlers) checked for CancellationToken parameter presence. Deterministic, adoption percentage.
Only 0/2 async methods accept a CancellationToken, so in-flight work can't be stopped early when the caller gives up — whatever ends it in your host (shutdown signal, timeout, abandoned request, user cancel). Thread a token through the call chain and honour it at each await and loop; where a method genuinely cannot be interrupted, omitting it is a deliberate choice — judge against your hosting model.
No CancellationToken parameter — this work can't be stopped early once started. (×2) — src/Cai.Web/Registry/RegistryEndpoints.cs:68, src/Cai.Web/Registry/RegistryEndpoints.cs:270
What to do
Thread a CancellationToken through async methods so work stops promptly on cancellation.
Other · Code Health — Whether an argument guard throws the exception its own condition describes — a guard that rejects a value for being EMPTY and reports it as `ArgumentNullException` tells the caller a parameter was null when it provably was not.
Method: Roslyn syntax: `throw new ArgumentNullException(nameof(p))` statements controlled by an `if`, whose condition is read for a test that is true of a NON-null `p` — an emptiness test that dereferences it (`p.Count == 0`, `!p.Any()`) or a BCL predicate documented true of the empty value (`string.IsNullOrEmpty(p)`). Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a `when` guard is free of side effects — a guard that increments a counter or assigns while deciding whether its arm matches applies that change during PATTERN MATCHING, on an arm that may not be selected, and skips it entirely when a short-circuit to its left answers first.
Method: Roslyn syntax: `when` guards on case labels and switch-expression arms, read for a mutation (`++`/`--`/assignment) sitting in a position the guard's own `&&`/`||`/`??`/`?:`/`?.` can skip. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a method that TAKES a lock or semaphore and gives it back from a flag-guarded `finally` returns the value that flag implies — reporting success while the guard hands the primitive back admits a second caller the exclusion was there to keep out, and reporting failure while the guard keeps it leaves nothing to ever give it back.
Method: Roslyn syntax: `try` statements whose `finally` releases a synchronisation primitive under a bare local-bool guard, where the method also TOOK that same primitive before the `try`, checked for a `return` of a bool literal whose value disagrees with the flag state the method's own straight-line assignments put it in. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether the work a diagnostic log line costs is paid only when that line is wanted — C# evaluates a call's arguments BEFORE the call, so a trace/debug message joined or projected out of a collection is built in full on every pass, and then discarded by a sink the shipped configuration leaves switched off.
Method: Roslyn syntax: log calls at a diagnostic level (a `Log`-prefixed method naming Trace/Debug/Verbose, or a bare `Debug`/`Trace`/`Verbose` on a receiver named for a logger), whose argument list is read for a call whose cost scales with a sequence — a LINQ operator, a materialisation, `string.Join`, a serializer — with no enclosing level check or conditional-compilation region. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a value the caller is invited to supply is the value the type actually uses — a constructor parameter stored in a private field that nothing ever reads while the default it was given is spelled out a second time at the site that should have read it, a keyed lookup that falls back to a different setting than the one its key names while the same type falls back to the matching one for that same key, or a culture-sensitive parse given no format provider by a type that feeds its own settable culture to the same kind of parse elsewhere. Either way, every caller who supplies a value silently gets something else.
Method: Roslyn syntax: private instance fields of a non-partial type assigned in a constructor from one of its own parameters with a `??` fallback, checked for whether anything in the type body reads the field and whether that same fallback expression is spelled out again outside the constructor; and `??` fallbacks onto a member access from a lookup call carrying exactly one string literal, grouped by that key across the type and checked for a fallback member whose folded name disagrees with the key while a sibling site for the same key agrees with it. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether a value handed from a callback to the body that waits on it crosses on something built to be crossed — a `Queue<T>`/`List<T>`/`Dictionary<K,V>` written inside an event handler and read back outside it is mutated by two flows at once, and the semaphore or completion source beside it orders how MANY items exist while leaving the collection's own head, tail and backing array unprotected.
Method: Roslyn syntax: method, accessor, local-function and lambda bodies that declare BOTH a non-thread-safe generic collection (`Queue`/`Stack`/`List`/`Dictionary`/`HashSet`/`Sorted*`/`LinkedList`) and a synchronisation primitive (`SemaphoreSlim`/`TaskCompletionSource`/`ManualResetEvent(Slim)`/`AutoResetEvent`/`CountdownEvent`) as locals, then read for a `+=`-registered lambda that raises that primitive while the body outside every lambda waits on it — and, in that scope, a mutating call on the collection inside the lambda paired with a mention of it outside. Any `lock` in the scope abstains it. Deterministic, provable per finding. Advisory.
Do you agree with this assessment?
X27 · Collection changed while being enumerated10.0 / 10Exemplary○ Nothing flagged
Other · Code Health — Whether a `foreach` leaves the collection it is walking alone — a body that adds to or removes from the very collection the loop is enumerating invalidates the enumerator it is holding, so the next `MoveNext` throws `InvalidOperationException` and the remaining items are never seen.
Method: Roslyn syntax + semantics: `foreach` statements whose body calls a structural mutator (`Add`/`Remove`/`Clear`/`Insert`/…) on the very expression the loop is enumerating. Two arms. ARM A — the source is a concrete fragile BCL collection, or a live `Keys`/`Values` view over one, and the mutator resolves to that same collection's own member; concurrent and immutable collections and arrays are outside the population by construction, since their enumerators survive a structural change. ARM B — the source is an argument-less accessor CALL on a receiver whose body is in source: the accessor must return a stored field VERBATIM and a sibling member must structurally change that same field, both read off the implementations rather than from the members' names. A mutation the loop provably exits immediately after (`break`/`return`/`throw`/`goto`), or one written inside a nested loop or a lambda, is counted and never reported. Deterministic, provable per finding. Advisory.
Do you agree with this assessment?
X28 · Index access outside its own emptiness guard10.0 / 10Exemplary○ Nothing flagged
Other · Code Health — Whether a condition that tests a value for emptiness indexes that same value only where the test holds — an `||` written one parenthesis too far to the left leaves an index access outside the guard beside it, so the empty case the guard exists to anticipate reaches the index and throws.
Method: Roslyn syntax only, no semantic model: the OUTERMOST `&&`/`||` of every boolean condition, read for a symbol the condition tests for emptiness (`string.IsNullOrEmpty`/`IsNullOrWhiteSpace`, a `Length`/`Count` comparison against a literal, `Any()`, a `Length`/`Count` pattern, or a comparison against `""`) and ALSO indexes. Each `symbol[...]` access is placed by a boolean-reachability walk from the access up to the outermost connective: an access is COVERED when some enclosing step has it in the right operand and the left operand, under the truth value that step forces, proves the symbol non-empty — a recursion over `&&`/`||` whose true- and false-directions are asymmetric. A finding needs BOTH an uncovered access and a covered one on the same symbol in the same condition, which is the agreeing twin that separates a misplaced parenthesis from an unrelated length test. Bare index accesses with no emptiness test in the condition are neither counted nor reported; a non-identifier receiver and a lambda nested inside the condition are outside the population. Deterministic, provable per finding. Advisory.
Do you agree with this assessment?
X29 · Per-element action decided by a fixed element10.0 / 10Exemplary○ Nothing flagged
Other · Code Health — Whether a decision taken once per element is taken ABOUT that element — a test inside a counted loop that reads a fixed subscript of the very collection its guarded statement indexes by the loop variable applies element zero's answer to all of them, so the elements that differ from it are all handled wrongly, and in the same direction.
Method: Roslyn syntax only, no semantic model: every `for` statement declaring exactly ONE loop variable, and every `if` inside its body that is not under a nested loop or a lambda. A site enters the population when the `if`’s condition never mentions the loop variable while the statement it guards indexes some collection by that variable ALONE (`c[i]`; `c[i + 1]` and `c[i, j]` are outside it). A finding additionally needs the AGREEING TWIN at the same-collection grain: the condition must read THAT SAME collection at a subscript that does not move — written into the condition, or reached through a local declared BEFORE the loop, so an alias bound inside the body is not followed. Both collection expressions must be simple identifiers. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether exceptions are handled rather than silently swallowed or rethrown with lost stack traces.
Method: Roslyn syntax scan: every catch clause counted; empty catches and bare rethrows flagged. Population is all catch clauses, not estimated. Deterministic, hard fact.
Do you agree with this assessment?
X30 · Support guard that admits what it rejects10.0 / 10Exemplary○ Nothing flagged
Other · Code Health — Whether a guard written as a NEGATED `||` says what its author meant — `!(a || b || x != k)` is `!a && !b && x == k` by De Morgan, so a bail-out that mixes capabilities the code needs with a fault it refuses turns inside out: it fires only where the capabilities are ABSENT, and lets every value the fault term names walk straight into the body that cannot handle it.
Method: Roslyn syntax only, no semantic model: every logical-not whose operand is a parenthesised `||` chain of two or more disjuncts, flattened (a left-nested `a || b || c` read once would see `(a || b)` as one disjunct). A site enters the population on that shape alone. A finding additionally needs the disjuncts to DISAGREE in polarity: at least one bare boolean read — an identifier or member access, never an invocation, which is a predicate rather than a capability flag — and at least one `x != <constant>`, the only form that negates into an exact-value pin (`== null` negates into a looser requirement and is outside the fault set). Consistently-polarised disjunctions, all-fault or all-capability, are counted and never reported; a negated `&&` is outside the population entirely. No same-receiver gate: it was measured to cost a real defect and remove no false positive. Deterministic, provable per finding. Advisory.
Do you agree with this assessment?
X32 · Type resolved by simple name across every loaded assembly10.0 / 10Exemplary○ Nothing flagged
Other · Code Health — Whether a plugin lookup names the type it means — searching every assembly loaded into the process for a candidate whose SIMPLE name equals a string supplied at runtime, and taking the first one found, is decided by assembly LOAD ORDER rather than by this source, so the same name can resolve to a different type on the next run.
Method: Roslyn syntax only, no semantic model: every invocation of `First`/`FirstOrDefault`/`Single`/`SingleOrDefault` whose OWN expression subtree contains both a `GetAssemblies()` call and a `GetTypes()`/`GetExportedTypes()` call — a single-element pick out of every type loaded into the process. A nested selector in the same chain sees no `GetAssemblies()` in its own subtree and is outside the population, so one lookup counts once however many links its chain has. A finding additionally needs both remaining halves: the selector must be `First`/`FirstOrDefault` (`Single`/`SingleOrDefault` reports the ambiguity rather than resolving it, and is counted and never reported), and the chain must carry an `==` comparison of `<lambda parameter>.Name` against something that is not a literal. The receiver must be a plain identifier bound by one of the chain’s own lambdas, which places `assembly.GetName().Name == "X"` outside the rule by construction. One exemption: a `.Name` test joined by `&&` to a `FullName`/`AssemblyQualifiedName` test on the same identifier is spared; joined by `||` it is not. Deterministic, provable per finding. Advisory.
Other · Code Health — Whether log calls use message templates (queryable) rather than interpolated strings.
Method: Roslyn syntax scan: every log call-site counted; interpolated-string first-argument violations flagged. Population is all log calls, not estimated. Deterministic.
Other · Code Health — Whether nullable reference types are enabled and not undermined by heavy `!` suppression.
Method: Roslyn compiler-options scan: NullableContextOptions per project; null-forgiving (!) suppression density per 1k syntax nodes. Deterministic, adoption plus suppression penalty.
~1.1 `!` suppressions per 1k syntax nodes — 61 suppression(s) across the 53929 syntax node(s) in code where nullable warnings are ENABLED, which is the only code a `!` can suppress anything in (a `!` under `#nullable disable` is inert and is not counted, and its file's nodes are not in the denominator). Each one tells the compiler to trust you about null, suppressing the very safety NRTs provide.
What to do
Enable <Nullable>enable</Nullable> across all projects and resolve warnings rather than suppressing with `!`.
Do you agree with this assessment?
WCAG coverage — what static analysis assessed
Statically assessed 10 of 55 WCAG 2.2 Level A/AA success criteria (18%; ≈20% of the 50 WCAG 2.1 AA criteria for EN 301 549). The other 45 require runtime or manual evaluation. Partial signal only (a clean result is necessary, not sufficient; static analysis fully verifies none). This is accessibility readiness, not a conformance claim — a WCAG conformance claim requires manual evaluation (WCAG-EM 1.0).
The score is the rank-weighted fold of these lenses (worst-heaviest), each including its meta-dimensions; a lens with a Critical contributor is capped at Fair (its band reads "gated by …") and is never the strongest area however high its average.
Not evidenced — 1 control(s) we could not find positive evidence for
These checks grade a working control, and the repository shows no evidence of one. That is deliberately not scored as a zero: a repository cannot show an ops runbook, a database TTL or an infrastructure-side audit log, so absence of evidence here is not evidence the control is missing. It is also not a statement that the check is irrelevant to this codebase — the thing it grades applies; we just could not see it. Excluded from the score either way.
P5 DR & Backup — not evidenced — repo shows no backup/RTO/RPO controls; absence of evidence is not evidence of a working control
Not included — 30 check(s) not relevant to this codebase
These checks had nothing to measure here (no tests, no git history, the codebase is small, or the architecture style doesn't apply), so they're omitted above rather than scored low.
AC1 Text alternatives — No image/media element found in the parsed markup — AC1 not applicable here.
AC4 Keyboard semantics — No interactive element found in the parsed markup — AC4 not applicable here.
AX7 Slice cohesion — not applicable — not a vertical-slice architecture
AX9 CQS / query purity — no CQRS query handlers detected — query purity is not applicable to this codebase
AXB2 Runtime readiness — Advisory — this card reports evidence and never carries a score, so there is nothing missing here.
C1 Data Protection — No personal data detected in the analyzed source — no PII-typed entity/column names (Email, FirstName, DateOfBirth, …), no ASP.NET Identity / user-account model, and no stored user credentials. GDPR data-protection controls are therefore N/A here. If this is intentional, record the no-PII posture in an ADR; if the app does process personal data, name those fields conventionally so this dimension activates.
C3 Audit Trail — No personal data detected in the analyzed source — no PII-typed entity/column names (Email, FirstName, DateOfBirth, …), no ASP.NET Identity / user-account model, and no stored user credentials. GDPR data-protection controls are therefore N/A here. If this is intentional, record the no-PII posture in an ADR; if the app does process personal data, name those fields conventionally so this dimension activates.
C4 Data Retention — No personal data detected in the analyzed source — no PII-typed entity/column names (Email, FirstName, DateOfBirth, …), no ASP.NET Identity / user-account model, and no stored user credentials. GDPR data-protection controls are therefore N/A here. If this is intentional, record the no-PII posture in an ADR; if the app does process personal data, name those fields conventionally so this dimension activates.
C5 Data-Subject Rights — No personal data detected in the analyzed source — no PII-typed entity/column names (Email, FirstName, DateOfBirth, …), no ASP.NET Identity / user-account model, and no stored user credentials. GDPR data-protection controls are therefore N/A here. If this is intentional, record the no-PII posture in an ADR; if the app does process personal data, name those fields conventionally so this dimension activates.
D16 Bus Factor — single-maintainer repository — bus factor is not applicable
D23 Boundary Type-Coupling — Cross-context type coupling could not be assessed — this codebase's bounded contexts are neither declared nor inferable.
D31 IaC & Container Security — No Infrastructure-as-Code or container manifests found (Dockerfile, Docker Compose, Terraform, Kubernetes/Helm, CloudFormation, ARM, Bicep, Ansible); nothing to scan.
D32 Data Compliance (PII/GDPR) — 12 file(s) were not parsed by semgrep — the PII/GDPR ruleset never ran over them
D37 Vulnerability-disclosure Policy — No vulnerability-disclosure policy file found (SECURITY.md/.markdown/.rst/.txt at root or under .github/.forgejo/.gitea/docs, .well-known/security.txt). A coordinated-disclosure policy may live off-repo, so this is not evidenced rather than failed.
D40 Network Egress Confinement — No Kubernetes/orchestration workloads found in the repository manifests; network egress policy is a cluster-native control that may live at the platform/firewall layer, so there is nothing to assess here.
D41 Kernel & Syscall Confinement — No Kubernetes/orchestration workloads found in the repository manifests; seccomp/AppArmor/SELinux confinement is a workload-level control, so there is nothing to assess here.
D42 Runtime Threat Enforcement — No Kubernetes/orchestration workloads found in the repository manifests; runtime threat-detection and admission-control policy are cluster-level controls, so there is nothing to assess here.
D8 Code Coverage — Coverage NOT MEASURED: `--collect:"XPlat Code Coverage"` names a data collector that ships in the `coverlet.collector` package, and this repository wires up none — no test project references it and no runsettings declares one. The absence of coverage here is therefore not evidence about the suite or about our analyzer environment: without a collector, `--collect` produces nothing even from a suite that builds and passes. Add a `coverlet.collector` PackageReference to the test project(s) (or commit the Cobertura/OpenCover/lcov report your CI produces) and real coverage will be measured. It is excluded from the score rather than counted as a near-zero defect.
DM1 Domain Modelling — not scored — this repository shows only 1 of the 3 signals this lens looks for (7 value object(s))
ED1 Event-Driven — not scored — this repository shows none of the 3 signals this lens looks for
ED5 Idempotency — no mutating command handlers or message consumers detected — idempotency check not applicable
ES1 Event Sourcing — not scored — this repository shows none of the 3 signals this lens looks for
P12 CI test-gate honesty — Reported, not scored — this card publishes what the CI gate does with the test inventory rather than grading it. The findings above are its output.
P8 Schema migrations — no EF Core usage detected
P9 Domain vs controller coverage — no coverage report found on disk — produce a coverage report in a standard format (Cobertura — `dotnet test --collect:"XPlat Code Coverage"` with a `coverlet.collector` PackageReference) and commit it — a hosted scan measures a clone of the repository, so a report that exists only in a working tree, a CI runner's or your own, never reaches it; the artefact is commonly gitignored, so `git add -f` that one file (or un-ignore its path) and commit it alongside the code it measures, or wire coverage collection into CI, to enable this cross-layer check
SC1 Supply-chain hygiene — Advisory — this card reports evidence and never carries a score, so there is nothing missing here.
X10 Duplicated predicate — Advisory — this card reports evidence and never carries a score, so there is nothing missing here.
X6 Hand-rolled structured-format parsing — Advisory — this card reports evidence and never carries a score, so there is nothing missing here.
X7 Silent fallback defaults — Advisory — this card reports evidence and never carries a score, so there is nothing missing here.
X9 Subsumed condition operand — Advisory — this card reports evidence and never carries a score, so there is nothing missing here.
Appendix A — Findings (grouped)
The findings behind the scores, grouped by severity, then by dimension and kind. The high-severity issues are enumerated in full below; items per group are capped at 25 with any overflow stated explicitly per group, never silently truncated. The complete machine-readable list of every finding (all severities) is the companion findings.md in this report's bundle.
High: dependabot-missing-cooldown .github/dependabot.yml:7— This Dependabot configuration does not set a cooldown period. Newly published packages can be malicious or unstable. Add a `cooldown` block with `default-days: 7` to each `package-ecosystem` entry under `updates` to wait 7 days before proposing updates to newly published package versions. Reference: https://docs.github.com/en/code-security/dependabot/dependabot-version-updates/configuration-options-for-the-dependabot.yml-file#cooldown. This is a semgrep security-AUDIT rule reporting a POLICY that is absent or weaker than its recommendation, not an exploitable defect. Confirm whether the current setting is a deliberate decision for this repository — and apply the change where it is not; where it is (a policy your release process already enforces elsewhere, or one this repository has consciously opted out of), record the decision and leave the configuration as it is. This configuration file has 2 such entries; one cooldown decision clears them all — reported once.
High: run-shell-injection .github/workflows/publish-packages.yml:41— Using variable interpolation `${{...}}` with `github` context data in a `run:` step could allow an attacker to inject their own code into the runner. This would allow them to steal secrets and code. `github` context data can have arbitrary user input and should be treated as untrusted. Instead, use an intermediate environment variable with `env:` to store the data and use the environment variable in the `run:` script. Reference it as a shell VARIABLE rather than a `${{ }}` interpolation, using your shell's own syntax (`"$ENVVAR"` in bash, `$env:ENVVAR` in PowerShell), so the value is passed as data and never re-expanded as code.
High: watchdog-untrusted-context-laundered-through-step-output .github/workflows/publish-packages.yml:54— This script interpolates a step output that carries an attacker-chosen value. The step that produced it, named in the same job, expands an untrusted workflow expression inside its own script and writes the result to `$GITHUB_OUTPUT` -- so the value leaves that step under a `steps.*.outputs.*` name while still being exactly the characters whoever triggered this workflow chose. Actions expands `${{ }}` before any shell exists, so those characters are written into THIS script as code rather than passed to it as data, and quoting the interpolation does not change that. A branch or tag name is not a safe token: Git's ref-name rules permit `$`, a backtick, `;`, `|` and `&`, and a dispatch input permits everything. Fixing only the step that first read the untrusted expression does not close this site -- the output still carries the value, and this script still executes it, in a job that holds this workflow's secrets. Bind the output to an `env:` entry on THIS step (`env:` `SAFE_VALUE: ${{ steps.the_step.outputs.the_value }}`) and read that ENVIRONMENT VARIABLE inside the script with your shell's own syntax (`"$SAFE_VALUE"` in bash, `$env:SAFE_VALUE` in PowerShell), so the value reaches the process as data and is never substituted into the script text. Validate the value against an explicit pattern in the producing step if its shape is known.
High: watchdog-untrusted-context-laundered-through-step-output .github/workflows/publish-packages.yml:67— This script interpolates a step output that carries an attacker-chosen value. The step that produced it, named in the same job, expands an untrusted workflow expression inside its own script and writes the result to `$GITHUB_OUTPUT` -- so the value leaves that step under a `steps.*.outputs.*` name while still being exactly the characters whoever triggered this workflow chose. Actions expands `${{ }}` before any shell exists, so those characters are written into THIS script as code rather than passed to it as data, and quoting the interpolation does not change that. A branch or tag name is not a safe token: Git's ref-name rules permit `$`, a backtick, `;`, `|` and `&`, and a dispatch input permits everything. Fixing only the step that first read the untrusted expression does not close this site -- the output still carries the value, and this script still executes it, in a job that holds this workflow's secrets. Bind the output to an `env:` entry on THIS step (`env:` `SAFE_VALUE: ${{ steps.the_step.outputs.the_value }}`) and read that ENVIRONMENT VARIABLE inside the script with your shell's own syntax (`"$SAFE_VALUE"` in bash, `$env:SAFE_VALUE` in PowerShell), so the value reaches the process as data and is never substituted into the script text. Validate the value against an explicit pattern in the producing step if its shape is known.
High: watchdog-secret-interpolated-into-run .github/workflows/publish-packages.yml:85— A secret is interpolated straight into this `run:` script. Actions expands `${{ }}` BEFORE the shell exists, so the literal secret value is what gets written into the step's script file on the runner's disk and passed to the shell as code — not as data. Quoting does not change that. From there it reaches every place the script itself reaches: the on-disk script, any shell trace (`set -x`, `bash -v`), a `ps` listing of the commands it launches, and any error text that echoes the failing command. Log masking does not cover those, and a secret cannot be un-leaked once it is on a runner you do not own. Bind the secret to an `env:` entry on the step instead (`env:` `MY_TOKEN: ${{ secrets.MY_TOKEN }}`) and reference that ENVIRONMENT VARIABLE from the script using your shell's own syntax (`"$MY_TOKEN"` in bash, `$env:MY_TOKEN` in PowerShell), so the value is passed to the process as data and is never substituted into the script text.
High: csharp-sqli src/Cai.Web/Noise/NoiseStore.cs:477— Detected a formatted string in a SQL statement. This could lead to SQL injection if variables in the SQL statement are not properly sanitized. Bind every value as a parameter instead of building the statement text: add a `DbParameter` to `DbCommand.Parameters` for each value and reference it by name in the SQL. This works the same way for every ADO.NET provider. This is a semgrep security-AUDIT rule: it reports that a sensitive construct is present, not that it is exploitable here. Confirm whether this site handles untrusted input or is reachable across a trust boundary — and apply the change where it is; where the construct is required by the platform or protocol it calls into, and carries no untrusted data (a syscall/FFI shim, a build- or debug-gated tool, a fixed local surface), record the review and leave the code as it is.
High: csharp-sqli src/Cai.Web/Noise/NoiseStore.cs:489— Detected a formatted string in a SQL statement. This could lead to SQL injection if variables in the SQL statement are not properly sanitized. Bind every value as a parameter instead of building the statement text: add a `DbParameter` to `DbCommand.Parameters` for each value and reference it by name in the SQL. This works the same way for every ADO.NET provider. This is a semgrep security-AUDIT rule: it reports that a sensitive construct is present, not that it is exploitable here. Confirm whether this site handles untrusted input or is reachable across a trust boundary — and apply the change where it is; where the construct is required by the platform or protocol it calls into, and carries no untrusted data (a syscall/FFI shim, a build- or debug-gated tool, a fixed local surface), record the review and leave the code as it is. Note what kind of statement this is: it is DDL — it creates, alters or drops a named object — and a DDL statement accepts no query parameters at ALL. Not the object's name, which no driver can bind because the statement must be planned before the parameter is known, and not the file paths, sizes or option values written beside it either, which are part of the statement's syntax rather than values in it. So "use a parameterised query" is not available for this site, and neither is any partial version of it. What IS available, and what this row is asking you to check: every name must be quoted with the dialect's own identifier quoter, and that quoter must DOUBLE the delimiter that would close the quoting — a helper that only wraps the value in brackets or quotes leaves the statement exactly where it started, because one embedded delimiter ends the name early and the rest is read as statement structure. If a quoting helper is already applied here, the review is whether it doubles; and every value in the statement that is NOT an identifier — a path, a size, a mode — needs to be checked against a fixed set of the values this code may emit before it is formatted in, because quoting does not constrain it. Where the values can only come from configuration you control, say so at the call site: that is the review this row is asking for.
High: csharp-sqli src/Cai.Web/Registry/RegistryStore.cs:214— Detected a formatted string in a SQL statement. This could lead to SQL injection if variables in the SQL statement are not properly sanitized. Bind every value as a parameter instead of building the statement text: add a `DbParameter` to `DbCommand.Parameters` for each value and reference it by name in the SQL. This works the same way for every ADO.NET provider. This is a semgrep security-AUDIT rule: it reports that a sensitive construct is present, not that it is exploitable here. Confirm whether this site handles untrusted input or is reachable across a trust boundary — and apply the change where it is; where the construct is required by the platform or protocol it calls into, and carries no untrusted data (a syscall/FFI shim, a build- or debug-gated tool, a fixed local surface), record the review and leave the code as it is.
High: csharp-sqli src/Cai.Web/Registry/RegistryStore.cs:226— Detected a formatted string in a SQL statement. This could lead to SQL injection if variables in the SQL statement are not properly sanitized. Bind every value as a parameter instead of building the statement text: add a `DbParameter` to `DbCommand.Parameters` for each value and reference it by name in the SQL. This works the same way for every ADO.NET provider. This is a semgrep security-AUDIT rule: it reports that a sensitive construct is present, not that it is exploitable here. Confirm whether this site handles untrusted input or is reachable across a trust boundary — and apply the change where it is; where the construct is required by the platform or protocol it calls into, and carries no untrusted data (a syscall/FFI shim, a build- or debug-gated tool, a fixed local surface), record the review and leave the code as it is. Note what kind of statement this is: it is DDL — it creates, alters or drops a named object — and a DDL statement accepts no query parameters at ALL. Not the object's name, which no driver can bind because the statement must be planned before the parameter is known, and not the file paths, sizes or option values written beside it either, which are part of the statement's syntax rather than values in it. So "use a parameterised query" is not available for this site, and neither is any partial version of it. What IS available, and what this row is asking you to check: every name must be quoted with the dialect's own identifier quoter, and that quoter must DOUBLE the delimiter that would close the quoting — a helper that only wraps the value in brackets or quotes leaves the statement exactly where it started, because one embedded delimiter ends the name early and the rest is read as statement structure. If a quoting helper is already applied here, the review is whether it doubles; and every value in the statement that is NOT an identifier — a path, a size, a mode — needs to be checked against a fixed set of the values this code may emit before it is formatted in, because quoting does not constrain it. Where the values can only come from configuration you control, say so at the call site: that is the review this row is asking for.
Secret: generic-api-key examples/cai-delivery.keys.json @ 098ef54:6 (committed 2026-08-07)— gitleaks matched rule 'generic-api-key' here, in git history. Deleting the file does not remove it — the commit persists on every clone, fork and backup — so the act is to ROTATE this credential and treat it as compromised. The "Rotate the exposed credentials" row carries the full procedure, including when a `.gitattributes` provenance declaration is the correct answer instead.
Secret: generic-api-key examples/cai-delivery.keys.json @ bc4880c:6 (committed 2026-07-05)— gitleaks matched rule 'generic-api-key' here, in git history. Deleting the file does not remove it — the commit persists on every clone, fork and backup — so the act is to ROTATE this credential and treat it as compromised. The "Rotate the exposed credentials" row carries the full procedure, including when a `.gitattributes` provenance declaration is the correct answer instead.
Secret: generic-api-key examples/cai-delivery.keys.json @ 2b10b91:6 (committed 2026-07-01)— gitleaks matched rule 'generic-api-key' here, in git history. Deleting the file does not remove it — the commit persists on every clone, fork and backup — so the act is to ROTATE this credential and treat it as compromised. The "Rotate the exposed credentials" row carries the full procedure, including when a `.gitattributes` provenance declaration is the correct answer instead.
Secret: generic-api-key examples/cai-delivery.sample-key.json @ 098ef54:4 (committed 2026-08-07)— gitleaks matched rule 'generic-api-key' here, in git history. Deleting the file does not remove it — the commit persists on every clone, fork and backup — so the act is to ROTATE this credential and treat it as compromised. The "Rotate the exposed credentials" row carries the full procedure, including when a `.gitattributes` provenance declaration is the correct answer instead.
Secret: generic-api-key examples/cai-delivery.sample-key.json @ 098ef54:5 (committed 2026-08-07)— gitleaks matched rule 'generic-api-key' here, in git history. Deleting the file does not remove it — the commit persists on every clone, fork and backup — so the act is to ROTATE this credential and treat it as compromised. The "Rotate the exposed credentials" row carries the full procedure, including when a `.gitattributes` provenance declaration is the correct answer instead.
Leaked secret: high-entropy-secret examples/cai-delivery.sample-key.json:5— high-entropy-secret detected. Treat the value as compromised: it is readable by everyone who has ever had the repository, and deleting the line does not un-publish it. In order — (1) REVOKE it at whatever issued it and issue a replacement, which is the only step that actually closes the exposure; (2) load the replacement at run time from your platform's secret store or the process environment instead of from the tree, so no future value is committable; (3) remove the file or line and add its path to the repository's ignore rules, so it cannot come back; (4) if the value was ever live, purge it from the history as well, since a clone taken before the deletion still carries it. If this is instead a FIXTURE — key material generated for tests and valid nowhere — then the exposure is nil and the fix is to make that legible: generate it in test setup, or keep it under a test-data path, so a reader (and this scan) can tell it from the real thing.
Hotspot: src/Cai.Web/Noise/NoiseStandardEndpoints.cs src/Cai.Web/Noise/NoiseStandardEndpoints.cs:50— src/Cai.Web/Noise/NoiseStandardEndpoints.cs changed 38 times in last 90 days, max cyclomatic complexity 133 in NoiseStandardEndpoints.MapNoiseStandard at line 50. Frequent change and high complexity in one file compound: schedule the next change to it to include carving out the part being edited, with the area under test before it moves. Counted over 2026-06-15..2026-09-13, the 90 days ending at the analysed commit. Reproduce with `git log --since='2026-06-15 15:21:50 +02:00' --until='2026-09-13 15:21:50 +02:00' --full-history --no-merges -- src/Cai.Web/Noise/NoiseStandardEndpoints.cs`: merges are excluded because a merge re-states changes already counted at their own commits, and history is NOT path-simplified because a change that reached the file through a merged branch is still a change to it. That command counts raw commits and can read HIGHER than this row, which counts a cherry-picked re-land, and a revert together with the commit it undoes, once each — a difference of several commits on a file whose history was re-landed or reverted inside the window.
Hotspot: src/Cai.Web/Noise/NoiseSubmissions.cs src/Cai.Web/Noise/NoiseSubmissions.cs:149— src/Cai.Web/Noise/NoiseSubmissions.cs changed 8 times in last 90 days, max cyclomatic complexity 30 in NoiseSubmissions.Accept at line 149. 1 of those changes was a fix/bug commit, and the other 7 changed it for other reasons — this file is under both repair and feature pressure. Before the next change lands here, make sure the area it touches is under test, then split that area out of the file so the following change is smaller than this one — a file this often edited pays the complexity back every time. Counted over 2026-06-15..2026-09-13, the 90 days ending at the analysed commit. Reproduce with `git log --since='2026-06-15 15:21:50 +02:00' --until='2026-09-13 15:21:50 +02:00' --full-history --no-merges -- src/Cai.Web/Noise/NoiseSubmissions.cs`: merges are excluded because a merge re-states changes already counted at their own commits, and history is NOT path-simplified because a change that reached the file through a merged branch is still a change to it. That command counts raw commits and can read HIGHER than this row, which counts a cherry-picked re-land, and a revert together with the commit it undoes, once each — a difference of several commits on a file whose history was re-landed or reverted inside the window.
Hotspot: src/Cai.Scoring/CaiScorer.cs src/Cai.Scoring/CaiScorer.cs:163— src/Cai.Scoring/CaiScorer.cs changed 9 times in last 90 days, max cyclomatic complexity 17 in CaiScorer.ScoreFromEvidence at line 163. 1 of those changes was a fix/bug commit, and the other 8 changed it for other reasons — this file is under both repair and feature pressure. Before the next change lands here, make sure the area it touches is under test, then split that area out of the file so the following change is smaller than this one — a file this often edited pays the complexity back every time. Counted over 2026-06-15..2026-09-13, the 90 days ending at the analysed commit. Reproduce with `git log --since='2026-06-15 15:21:50 +02:00' --until='2026-09-13 15:21:50 +02:00' --full-history --no-merges -- src/Cai.Scoring/CaiScorer.cs`: merges are excluded because a merge re-states changes already counted at their own commits, and history is NOT path-simplified because a change that reached the file through a merged branch is still a change to it. That command counts raw commits and can read HIGHER than this row, which counts a cherry-picked re-land, and a revert together with the commit it undoes, once each — a difference of several commits on a file whose history was re-landed or reverted inside the window.
Hotspot: src/Cai.Web/Noise/PublicationContract.cs src/Cai.Web/Noise/PublicationContract.cs:97— src/Cai.Web/Noise/PublicationContract.cs changed 5 times in last 90 days, max cyclomatic complexity 30 in PublicationContract.Check at line 97. 1 of those changes was a fix/bug commit, and the other 4 changed it for other reasons — this file is under both repair and feature pressure. Before the next change lands here, make sure the area it touches is under test, then split that area out of the file so the following change is smaller than this one — a file this often edited pays the complexity back every time. Counted over 2026-06-15..2026-09-13, the 90 days ending at the analysed commit. Reproduce with `git log --since='2026-06-15 15:21:50 +02:00' --until='2026-09-13 15:21:50 +02:00' --full-history --no-merges -- src/Cai.Web/Noise/PublicationContract.cs`: merges are excluded because a merge re-states changes already counted at their own commits, and history is NOT path-simplified because a change that reached the file through a merged branch is still a change to it. That command counts raw commits and can read HIGHER than this row, which counts a cherry-picked re-land, and a revert together with the commit it undoes, once each — a difference of several commits on a file whose history was re-landed or reverted inside the window.
Hotspot: src/Cai.Web/Registry/RegistryEndpoints.cs src/Cai.Web/Registry/RegistryEndpoints.cs:68— src/Cai.Web/Registry/RegistryEndpoints.cs changed 3 times in last 90 days, max cyclomatic complexity 15 in RegistryEndpoints.PublishAsync at line 68. 1 of those changes was a fix/bug commit, and the other 2 changed it for other reasons — this file is under both repair and feature pressure. Before the next change lands here, make sure the area it touches is under test, then split that area out of the file so the following change is smaller than this one — a file this often edited pays the complexity back every time. Counted over 2026-06-15..2026-09-13, the 90 days ending at the analysed commit. Reproduce with `git log --since='2026-06-15 15:21:50 +02:00' --until='2026-09-13 15:21:50 +02:00' --full-history --no-merges -- src/Cai.Web/Registry/RegistryEndpoints.cs`: merges are excluded because a merge re-states changes already counted at their own commits, and history is NOT path-simplified because a change that reached the file through a merged branch is still a change to it. That command counts raw commits and can read HIGHER than this row, which counts a cherry-picked re-land, and a revert together with the commit it undoes, once each — a difference of several commits on a file whose history was re-landed or reverted inside the window.
Medium: watchdog-eol-node-runtime-in-workflow .github/workflows/a11y.yml:29— This workflow installs a Node.js major version that has reached END OF LIFE, so every job in it - including, where this is a release or publish workflow, the job that produces the artefact users download - runs on an interpreter that no longer receives security patches. Node.js 16 left support on 2023-09-11, 18 on 2025-04-30, 20 on 2026-04-30, and 12, 14, 17, 19 and 21 earlier still; only 22 and 24 remain in support. An EOL runtime does not merely age - known vulnerabilities in V8, in the bundled OpenSSL, in the HTTP parser and in npm ship with it permanently, and no upgrade of the dependencies installed on top of it can reach them. Where the workflow builds a published artefact, the exposure is not confined to the runner: the artefact is compiled, bundled and packed by that interpreter and by the toolchain versions still compatible with it. Raise this value to a Node.js release that is still in support - 22 or 24 at the time of writing - and if the project genuinely must build on an older line, say so explicitly in the workflow rather than leaving the number to read as current. Note that pinning the action itself to an immutable revision does not address this at all; the version installed is this input, not the action. This is a semgrep security-AUDIT rule reporting a POLICY that is absent or weaker than its recommendation, not an exploitable defect. Confirm whether the current setting is a deliberate decision for this repository — and apply the change where it is not; where it is (a policy your release process already enforces elsewhere, or one this repository has consciously opted out of), record the decision and leave the configuration as it is.
Medium: unsafe-path-combine src/Cai.Scoring/RubricCatalogStore.cs:124— String argument rubricVersion is used to read or write data from a file via Path.Combine without direct sanitization via Path.GetFileName. If the path is user-supplied data this can lead to path traversal.
Medium: unsafe-path-combine src/Cai.Scoring/RubricCatalogStore.cs:150— String argument rubricVersion is used to read or write data from a file via Path.Combine without direct sanitization via Path.GetFileName. If the path is user-supplied data this can lead to path traversal.
MethodTooLong: NoiseStandardEndpoints.MapNoiseStandard src/Cai.Web/Noise/NoiseStandardEndpoints.cs:50— MethodTooLong — MapNoiseStandard runs 1338 significant lines (blank, comment-only and punctuation-only lines excluded) in one body. The bar is 100 significant lines; this is 1238 over it, 13.38× the bar. This is length, not branching: a long straight-line body scores low on complexity and is still read whole to change any part of it, so the complexity numbers beside this row neither confirm nor excuse it. To reduce it, extract each cohesive step of the body — the runs of statements that work on the same values and would earn the same name — into its own named unit, and have this one call them in order.
MethodTooLong: PublicationContract.Check src/Cai.Web/Noise/PublicationContract.cs:97— MethodTooLong — Check runs 138 significant lines (blank, comment-only and punctuation-only lines excluded) in one body. The bar is 100 significant lines; this is 38 over it, 1.38× the bar. This is length, not branching: a long straight-line body scores low on complexity and is still read whole to change any part of it, so the complexity numbers beside this row neither confirm nor excuse it. To reduce it, extract each cohesive step of the body — the runs of statements that work on the same values and would earn the same name — into its own named unit, and have this one call them in order.
MethodTooLong: NoiseSubmissions.Accept src/Cai.Web/Noise/NoiseSubmissions.cs:149— MethodTooLong — Accept runs 127 significant lines (blank, comment-only and punctuation-only lines excluded) in one body. The bar is 100 significant lines; this is 27 over it, 1.27× the bar. This is length, not branching: a long straight-line body scores low on complexity and is still read whole to change any part of it, so the complexity numbers beside this row neither confirm nor excuse it. To reduce it, extract each cohesive step of the body — the runs of statements that work on the same values and would earn the same name — into its own named unit, and have this one call them in order.
D10 · Test Quality· Depends on a live external host · ×2
Depends on a live external host: A_marketing_origin_may_call_the_api_from_a_browser tests/Cai.Tests/PublicCorsTests.cs:28— This test fetches `https://codeassuranceindex.info` while it runs, so it passes only while codeassuranceindex.info is up, reachable from the build machine, and still serving what this test expects — none of which this repository controls. It goes red on an aeroplane, behind a corporate proxy, in a sealed CI network, and on the day the third party edits that document. Serve the payload from the repository instead (a checked-in fixture file), or stub the transport; if the point really is to check the remote contract, move it out of the unit suite into a separately-scheduled integration job that is allowed to fail for someone else's reasons.
Depends on a live external host: Credentials_are_never_allowed tests/Cai.Tests/PublicCorsTests.cs:67— This test fetches `https://codeassuranceindex.info` while it runs, so it passes only while codeassuranceindex.info is up, reachable from the build machine, and still serving what this test expects — none of which this repository controls. It goes red on an aeroplane, behind a corporate proxy, in a sealed CI network, and on the day the third party edits that document. Serve the payload from the repository instead (a checked-in fixture file), or stub the transport; if the point really is to check the remote contract, move it out of the unit suite into a separately-scheduled integration job that is allowed to fail for someone else's reasons.
FileTooLong: Noise/NoiseStandardEndpoints.cs src/Cai.Web/Noise/NoiseStandardEndpoints.cs— FileTooLong — 1618 significant lines (blank, comment-only and punctuation-only lines excluded). The bar is 500 significant lines; this is 1118 over it, 3.24× the bar. To reduce it, split the file along the responsibilities already in it: move each cohesive group of declarations into its own sibling file in the same module or package, so no one file has to be read whole to change one of them.
FileTooLong: Noise/NoiseStore.cs src/Cai.Web/Noise/NoiseStore.cs— FileTooLong — 557 significant lines (blank, comment-only and punctuation-only lines excluded). The bar is 500 significant lines; this is 57 over it, 1.11× the bar. To reduce it, split the file along the responsibilities already in it: move each cohesive group of declarations into its own sibling file in the same module or package, so no one file has to be read whole to change one of them.
AC2 · Forms & labels· <fieldset> without a <legend> · ×1
<fieldset> without a <legend> src/Cai.Web/Components/Pages/NoiseRate.razor:183— A fieldset groups related controls but has no <legend> to name the group. Add a <legend> as its first child — or, if the group already has a visible caption beside it (or the fieldset is there only to disable its subtree and carries no group box), point aria-labelledby at that caption's id instead, which names the group without rendering a second one.
NoiseStandardEndpoints.MapNoiseStandard (cyclomatic 133) src/Cai.Web/Noise/NoiseStandardEndpoints.cs:50— NoiseStandardEndpoints.MapNoiseStandard has cyclomatic complexity 133 (threshold 15). Most of this is not in the body itself: 1 of the 133 points is its own statement and the rest belongs to 36 function literals inside it that branch (lines 1780, 860, 1181, …). The decisions are inside those literals, which nothing outside this body can call, review or test on its own, so splitting the enclosing body is not the move available here. To reduce it, lift the literals' work into a named function or method at the enclosing scope and have each literal call it, then reduce whichever part then reads as the largest.
NoiseSubmissions.Accept (cyclomatic 30) src/Cai.Web/Noise/NoiseSubmissions.cs:149— NoiseSubmissions.Accept has cyclomatic complexity 30 (threshold 15). To reduce it, separate the cases: extract each independent branch into its own named function, and where the body has guards that only reject input, fold those into early returns at the top.
PublicationContract.Check (cyclomatic 30) src/Cai.Web/Noise/PublicationContract.cs:97— PublicationContract.Check has cyclomatic complexity 30 (threshold 15). To reduce it, separate the cases: extract each independent branch into its own named function, and where the body has guards that only reject input, fold those into early returns at the top.
CaiScorer.ScoreFromEvidence (cyclomatic 17) src/Cai.Scoring/CaiScorer.cs:163— CaiScorer.ScoreFromEvidence has cyclomatic complexity 17 (threshold 15). Of this number, 16 points are the body's own statements and 1 belongs to one function literal inside it that branches. To reduce it, separate the cases: extract each independent branch into its own named function, and where the body has guards that only reject input, fold those into early returns at the top.
Repeated repair: src/Cai.Web/ApiRateLimiting.cs src/Cai.Web/ApiRateLimiting.cs:188— src/Cai.Web/ApiRateLimiting.cs changed 5 times in last 90 days and 3 of those changes were fix/bug commits, so repair is the majority of this file's churn. Its max cyclomatic complexity is 13 (its worst body is ApiRateLimiting.Compute at line 188), UNDER the 15 threshold, so this is deliberately not filed as a churn × complexity hotspot — the difficulty here is in the behaviour the file has to get right, not in its control flow, and refactoring it for complexity would be the wrong move. The repairs counted were: “fix(web): the app host's own redirect table still pointed home”; “fix(web): the standard had two websites, and the one with the working tools was unreachable”; “fix(web): traffic-class rate limiter — registry principals and public probes off the anonymous budget”. Each one is a case this code did not handle. Before the next change lands here, check that every one of them is pinned by a test that fails without its fix; where the same area keeps coming back, the durable fix is usually at the interface that keeps being misused rather than at the line that was last corrected. Counted over 2026-06-15..2026-09-13, the 90 days ending at the analysed commit. Reproduce with `git log --since='2026-06-15 15:21:50 +02:00' --until='2026-09-13 15:21:50 +02:00' --full-history --no-merges -- src/Cai.Web/ApiRateLimiting.cs`: merges are excluded because a merge re-states changes already counted at their own commits, and history is NOT path-simplified because a change that reached the file through a merged branch is still a change to it. That command counts raw commits and can read HIGHER than this row, which counts a cherry-picked re-land, and a revert together with the commit it undoes, once each — a difference of several commits on a file whose history was re-landed or reverted inside the window.
NoiseStandardEndpoints.MapNoiseStandard (cognitive 151) src/Cai.Web/Noise/NoiseStandardEndpoints.cs:50— NoiseStandardEndpoints.MapNoiseStandard has cognitive complexity 151 (threshold 15). Most of this is not in the body itself: 0 of the 151 points are its own statements and the rest belongs to 36 function literals inside it that branch (lines 1780, 860, 1181, …). The decisions are inside those literals, which nothing outside this body can call, review or test on its own, so splitting the enclosing body is not the move available here. To reduce it, lift the literals' work into a named function or method at the enclosing scope and have each literal call it, then reduce whichever part then reads as the largest.
PublicationContract.Check (cognitive 43) src/Cai.Web/Noise/PublicationContract.cs:97— PublicationContract.Check has cognitive complexity 43 (threshold 15). To reduce it, flatten the nesting: invert conditions into early returns or guard clauses so the happy path stays at one level, and lift the deepest nested block into its own named function.
NoiseSubmissions.Accept (cognitive 42) src/Cai.Web/Noise/NoiseSubmissions.cs:149— NoiseSubmissions.Accept has cognitive complexity 42 (threshold 15). To reduce it, split the body into named stages: move each independent step or branch into its own named function so the body reads as a short sequence of named calls rather than one long body.
CaiScorer.ScoreFromEvidence (cognitive 25) src/Cai.Scoring/CaiScorer.cs:163— CaiScorer.ScoreFromEvidence has cognitive complexity 25 (threshold 15). Of this number, 24 points are the body's own statements and 1 belongs to one function literal inside it that branches. To reduce it, split the body into named stages: move each independent step or branch into its own named function so the body reads as a short sequence of named calls rather than one long body.
Rejudge.Compare (cognitive 16) src/Cai.Web/Noise/Rejudge.cs:136— Rejudge.Compare has cognitive complexity 16 (threshold 15). To reduce it, split the body into named stages: move each independent step or branch into its own named function so the body reads as a short sequence of named calls rather than one long body.
D22 · Internal API Consistency· Duplicate intent · ×1
Duplicate intent: Both `Bands` and `BandCutlines` provide a `For` method that maps a score to a `Band`. `BandCutlines` is a property of `ScoringParameters` and likely contains the specific cutlines used for a version, while `Bands` appears to be a static utility. This creates ambiguity on which class to use for score-to-band conversion. — Remove `Bands.For` and rely exclusively on `BandCutlines.For` (or `ScoringParameters.Bands.For`) to ensure the banding logic is tied to the specific rubric version's parameters. (signatures: Cai.Scoring.Bands.For(double scoreZeroToOneHundred) | Cai.Scoring.BandCutlines.For(double scoreZeroToOneHundred))
D22 · Internal API Consistency· Confusingly similar names · ×1
Confusingly similar names: `Sign` and `SignPackage` are distinct operations (one returns `DeliverySignature`, the other `DeliveryPackage`), but the naming convention suggests `SignPackage` might be a variant of `Sign` rather than a distinct higher-level operation. It is unclear if `Sign` is a low-level primitive or if `SignPackage` is the primary entry point. — Rename `Sign` to `CreateSignature` or `SignPayload` to clearly distinguish it from `SignPackage`, which implies the creation of the final signed artifact. (signatures: Cai.Delivery.DeliverySigner.Sign(DeliveryPayload payload) | Cai.Delivery.DeliverySigner.SignPackage(DeliveryPayload payload))
D22 · Internal API Consistency· Ambiguous boolean semantics · ×1
Ambiguous boolean semantics: `SignatureValid` checks cryptographic integrity, while `AuthenticAndReproducing` likely checks semantic correctness (score reproduction). However, `Reproduced` is also a property. The naming `AuthenticAndReproducing` is a compound boolean that obscures whether it is a logical AND of `SignatureValid` and `Reproduced` or a separate check. If it is derived, it should be a method or clearly documented as a composite property. — If `AuthenticAndReproducing` is simply `SignatureValid && Reproduced`, expose it as a computed property with a clear name like `IsVerified` or remove it in favor of checking the two constituent properties explicitly. (signatures: Cai.Delivery.DeliveryVerification.AuthenticAndReproducing | Cai.Delivery.DeliveryVerification.SignatureValid)
ClassTooLong: NoiseStandardEndpoints src/Cai.Web/Noise/NoiseStandardEndpoints.cs— ClassTooLong — 1551 significant lines (blank, comment-only and punctuation-only lines excluded), 19 methods. The bar is 400 significant lines; this is 1151 over it, 3.88× the bar. The type holds no instance state, so there is no shared data to group its members by. To reduce it, split it by area instead: give each cohesive family of members its own smaller type, so no one type has to be read whole to change one of them.
TooManyMethods: SqliteNoiseStore src/Cai.Web/Noise/NoiseStore.cs— TooManyMethods — 498 significant lines (blank, comment-only and punctuation-only lines excluded), 34 methods. The bar is 30 methods; this is 4 over it, 1.13× the bar. To reduce it, group the members that share the same data into a smaller type of their own and delegate to it, so no single type carries every responsibility.
Medium CVE: Microsoft.Build.Tasks.Git 10.0.300 — Microsoft.Build.Tasks.Git 10.0.300 (transitive) has a Medium advisory; affects 2 projects — one upgrade fixes all. https://github.com/advisories/[GHSA redacted]
D36 · Supply-chain Provenance & Signing· Secret passed as a command-line argument · ×1
Secret passed as a command-line argument — 2 CI command(s) pass a credential in the command line (as an argument of its own, or attached to an option as `--flag=<secret>`), where it is visible in the runner's process table to any other process on the host (and to anything that logs a command line): .github/workflows/publish-packages.yml: dotnet nuget push "nupkg/*.nupkg" --source "https://api.nuget.org/v3/index.json" --api-key "$NUGET_API_KEY" --skip-duplicate; .github/workflows/publish-packages.yml: --api-key ${{ secrets.GITHUB_TOKEN }}. Pass the credential through the environment instead (an `env:` mapping on the step, read by the tool from its own variable) or on stdin, so it never appears in an argument vector.
Duplicated block (18 lines × 2) src/Cai.Web/Noise/NoiseStore.cs:474— src/Cai.Web/Noise/NoiseStore.cs:474-491 | src/Cai.Web/Registry/RegistryStore.cs:211-228 — the copies span different directories, so extracting a shared function means choosing where it lives: put it somewhere both call sites can already reach — a location they all depend on today, or a new shared one if there is none — and call it from each site; until then, every change has to be made twice.
Duplicated block (12 lines × 2) src/Cai.Delivery/Keys.cs:13— src/Cai.Delivery/Keys.cs:13-24 | src/Cai.Delivery/Keys.cs:62-73 — both copies are in the same file, so extract the block into one function there and call it from each site — the copies drift apart the first time only one of them is edited. Read the line range as the matched WINDOW rather than a finished unit: at `src/Cai.Delivery/Keys.cs:13` it does not close everything it opens, so those exact lines cannot be lifted as they stand — widen the region to the smallest complete statement or declaration that contains it, and extract that.
Duplicated block (6 lines × 2) src/Cai.Web/Noise/NoiseStore.cs:282— src/Cai.Web/Noise/NoiseStore.cs:282-287 | src/Cai.Web/Registry/RegistryStore.cs:131-136 — the copies span different directories, so extracting a shared function means choosing where it lives: put it somewhere both call sites can already reach — a location they all depend on today, or a new shared one if there is none — and call it from each site; until then, every change has to be made twice. Note first that the copies are not typed on the same thing: the declarations holding them bind `logger` to `ILogger<SqliteNoiseStore>` in one and `ILogger<SqliteRegistryStore>` in another, and the duplicated lines use it. The extracted unit therefore needs a parameter type that fits BOTH — their common supertype where they have one, or a new abstraction over them where they do not — and settling that is the step that comes BEFORE the extraction above. Where the two types are deliberately unrelated, the duplication is the price of that separation and the honest resolution is to record the decision rather than to extract.
Off the main sequence: Cai.Scoring — Cai.Scoring: abstractness 0.00, instability 0.00, distance 1.00 — zone of pain — concrete and depended on by 3 project(s), so it's rigid to change.
Coverage not measured — no coverage collector is wired up — Coverage NOT MEASURED: `--collect:"XPlat Code Coverage"` names a data collector that ships in the `coverlet.collector` package, and this repository wires up none — no test project references it and no runsettings declares one. The absence of coverage here is therefore not evidence about the suite or about our analyzer environment: without a collector, `--collect` produces nothing even from a suite that builds and passes. Add a `coverlet.collector` PackageReference to the test project(s) (or commit the Cobertura/OpenCover/lcov report your CI produces) and real coverage will be measured. It is excluded from the score rather than counted as a near-zero defect.
P12 · CI test-gate honesty· Test suite runs only after the merge · ×1
Test suite runs only after the merge — `deploy.yml`, `publish-packages.yml` run(s) the test suite, but no workflow that runs tests is triggered by a pull request (or a merge queue) — so the suite reports on code that is already on the default branch. A red build there blocks nothing and the only remedy is a revert. Add the pull-request trigger to the workflow that runs your suite so the gate applies before the merge, not after it.
SC1 · Supply-chain hygiene· NuGet dependencies are not locked · ×1
NuGet dependencies are not locked — No packages.lock.json and no central package management — restores aren't reproducible or pinned (SSDF PW.4.4). Enable <RestorePackagesWithLockFile>true</RestorePackagesWithLockFile> (commit the lockfile) or adopt Directory.Packages.props. Advisory — never scored.
Low: request-host-used deploy/nginx/api.codeassuranceindex.info.conf:16— '$http_host' and '$host' variables may contain a malicious value from attacker controlled 'Host' request header. Use an explicitly configured host value or a allow list for validation. This is a semgrep security-AUDIT rule reporting a POLICY that is absent or weaker than its recommendation, not an exploitable defect. Confirm whether the current setting is a deliberate decision for this repository — and apply the change where it is not; where it is (a policy your release process already enforces elsewhere, or one this repository has consciously opted out of), record the decision and leave the configuration as it is.
Low: request-host-used deploy/nginx/app.codeassuranceindex.info.conf:21— '$http_host' and '$host' variables may contain a malicious value from attacker controlled 'Host' request header. Use an explicitly configured host value or a allow list for validation. This is a semgrep security-AUDIT rule reporting a POLICY that is absent or weaker than its recommendation, not an exploitable defect. Confirm whether the current setting is a deliberate decision for this repository — and apply the change where it is not; where it is (a policy your release process already enforces elsewhere, or one this repository has consciously opted out of), record the decision and leave the configuration as it is.
Low: request-host-used deploy/nginx/codeassuranceindex.info.conf:61— '$http_host' and '$host' variables may contain a malicious value from attacker controlled 'Host' request header. Use an explicitly configured host value or a allow list for validation. This is a semgrep security-AUDIT rule reporting a POLICY that is absent or weaker than its recommendation, not an exploitable defect. Confirm whether the current setting is a deliberate decision for this repository — and apply the change where it is not; where it is (a policy your release process already enforces elsewhere, or one this repository has consciously opted out of), record the decision and leave the configuration as it is.
Low: request-host-used deploy/nginx/codeassuranceindex.info.conf:62— '$http_host' and '$host' variables may contain a malicious value from attacker controlled 'Host' request header. Use an explicitly configured host value or a allow list for validation. This is a semgrep security-AUDIT rule reporting a POLICY that is absent or weaker than its recommendation, not an exploitable defect. Confirm whether the current setting is a deliberate decision for this repository — and apply the change where it is not; where it is (a policy your release process already enforces elsewhere, or one this repository has consciously opted out of), record the decision and leave the configuration as it is.
Documentation: contradicts the code docs/CHALLENGE.md— The document describes the v1 registry API, but the companion cai-delivery package format (v1 addendum) is now out-of-date and no longer referenced in the text. Update to reflect that the v2 delivery package format is implemented and its companion endpoints are live.
Documentation: written for insiders docs/brand/README.md— The brand section references `cai-mark.svg`, `cai-favicon-64.png`, and `cai-og.png` as inline assets but does not explain the mark's purpose or how it relates to the CAI standard. Add a sentence stating what the band/glass callout signifies (the reading magnified) so readers understand its role in conveying the standard without brand colour.
Documentation: no project overview — The repository's README files do not describe what CAI is, its purpose, or its main value proposition. Add a one-line summary stating that CAI is an open reproducible scoring standard for codebase assurance.
Fat interface: INoiseStore (28 members) src/Cai.Web/Noise/NoiseStore.cs:79— `INoiseStore` declares 28 members: `TryRecordSubmission`, `FindSubmission`, `AlreadySubmitted`, `ListSubmissions`, `ConfigurationJson`, `RecordVerdict`, `RecordResolution`, `RegisterPrompt`, `ListVerdicts`, `ListResolutions`, `ListPrompts`, `RecordPublication`, `LatestPublication`, `PublishedPeriods`, `PublishedTallies`, `RecordRejudge`, `ListRejudge`, `RaiseDispute`, `FindDispute`, `ResolveDispute`, `ListDisputes`, `RegisterIntent`, `ListIntent`, `RecordFindings`, `FindFinding`, `RecordCost`, `CostFor`, `JudgedPeriods`. Counted as the author wrote them — a property is ONE member and its get/set accessors are not counted separately, and an event counts once. A wide interface forces every implementer and caller to depend on methods they don't use (the Interface-Segregation 'I' in SOLID). Split it into focused role-interfaces.
Anonymous-heavy surface — [AllowAnonymous] (15) outweighs [Authorize] (1) — review whether the open surface is intended.
D21 · Naming Consistency· The property `Ct` is used in two different test classes. While `Ct` is a common abbreviation for 'Context' or 'Count' in testing frameworks, using the same short, ambiguous abbreviation for potentially different concepts (e.g., a test context object vs. a count of items) reduces readability. It is preferable to use descriptive names like `TestContext` or `ItemCount`. · ×1
The property `Ct` is used in two different test classes. While `Ct` is a common abbreviation for 'Context' or 'Count' in testing frameworks, using the same short, ambiguous abbreviation for potentially different concepts (e.g., a test context object vs. a count of items) reduces readability. It is preferable to use descriptive names like `TestContext` or `ItemCount`. — Rename to descriptive names such as `TestContext` or `Count` depending on the actual type and usage. (symbols: Cai.Tests.VocabularyDriftTests.Ct, Cai.Tests.JudgePanelShapeTests.Ct)
D26 · Project Cohesion· Projects may be oversized for their cohesion · ×1
Projects may be oversized for their cohesion — 1 of 6 project(s) overshoot their size bounds, lowering Project Cohesion to 7.1/10. The most over is `Cai.Web` (10165 LoC, 139 public types across 4 namespaces). Review these for cohesion — split a project that spans unrelated responsibilities.
D28 · Secrets (history)· Rotate the exposed credentials · ×1
Rotate the exposed credentials — git history can't be un-committed — Some of these secrets are in git HISTORY: deleting the file does not remove them (the commit persists on every clone, fork and backup). The remediation is to ROTATE each historically-exposed credential and treat it as compromised — not to delete the file. Rewriting history is disruptive and unreliable across existing forks. (Working-tree-only secrets — no commit — can instead be removed from the file and moved to a secret store.) If flagged rows sit in content that is not this repository's own authored code — tool-emitted artifacts, recorded scanner output, wholesale-vendored third-party files — declare that provenance in .gitattributes (`path/** linguist-generated` or `linguist-vendored`): this scanner honors the repository's own declarations and stops counting such content, history-anchored rows included, on the next scan. The declaration is a reviewable statement about provenance, not a suppression — never use it on files that carry your own credentials. Every location above sits inside a test/fixture/sample tree, so there may be no live credential to revoke — in that case the performable actions are different ones: confirm each value was never reused outside the tests (a fixture key shared with a staging or demo environment IS a live credential and must be rotated), generate this material at test time instead of committing it so the next one cannot be mistaken for a real leak, and record the deliberate exposure where a reader of the file will see it. Rotate anything that fails the first check.
D36 · Supply-chain Provenance & Signing· Release publish has no approval gate · ×1
Release publish has no approval gate — 1 workflow(s) publish a package to a registry with no point at which a person is asked to approve the run (publish-packages.yml:79). Reached by a release tag any person or credential with write access to the repository can create — the filter selects release tags, so it is not reached by any tag at all, but it is reached by anyone who can push one of that shape. In every case the run is started by ONE actor with write access, acting alone — pushing a tag or clicking Run workflow is a deliberate act, but deliberate is not the same control as approved, and merging a pull request is not even deliberate. An `environment:` is the only Actions mechanism that can require a SECOND, named person. Each of these workflows also injects at least one long-lived named credential that every job in it can read — often, but not always, the registry key itself. What makes this worth the gate is that the artifact is not recallable: once a version is accepted by a public registry it has been fetched, cached by mirrors and pinned into other people's lockfiles within minutes, and the registries that matter refuse a re-upload of that version with different content, so a mistaken or hostile publish is corrected only by shipping another version and asking everyone downstream to move. Give the publishing job an `environment:` and configure that environment with required reviewers, then hold the registry credential on the environment rather than leaving it readable by every job in the repository — that is what turns the publish into an action a named person takes. Where the pipeline can use the registry's OIDC trusted publishing instead (`id-token: write`), that removes the standing credential altogether.
IL efficiency: 8 authored method(s) exceed the IL budget src/Cai.Web/Noise/NoiseSubmissions.cs:153— 8 of 443 first-party methods compile to oversized IL bodies (> 250 instructions); worst: Cai.Web.Noise.NoiseSubmissions.Accept @ src/Cai.Web/Noise/NoiseSubmissions.cs:153, 685 IL instructions; that pulled this dimension to 9.6/10. These bodies are far past the JIT's inline budget, so splitting them does not make them inlinable — what moves the number is emitting less: collapsing LINQ chains and closures on hot paths, and interpolation built eagerly where it is only sometimes used.
P10 · Library API & versioning· Large public API surface · ×1
Large public API surface — 187/207 types (90%) are public. For a library, every public type is a stability contract — make internal-by-default and expose only the intended API.
Logging is not universal — Only 2/3 service-like projects use logging (pure contract/DTO projects are excluded — they have nothing to log). Of those 3, 2 ship a process this repository operates; the rest are libraries their consumer hosts, where the logging decision belongs to the host.
X2 · Cancellation propagation· Not all async methods take a CancellationToken · ×1
Not all async methods take a CancellationToken — Only 0/2 async methods accept a CancellationToken, so in-flight work can't be stopped early when the caller gives up — whatever ends it in your host (shutdown signal, timeout, abandoned request, user cancel). Thread a token through the call chain and honour it at each await and loop; where a method genuinely cannot be interrupted, omitting it is a deliberate choice — judge against your hosting model.
Null-forgiving operator (`!`) suppressions reduce the NRT score — ~1.1 `!` suppressions per 1k syntax nodes — 61 suppression(s) across the 53929 syntax node(s) in code where nullable warnings are ENABLED, which is the only code a `!` can suppress anything in (a `!` under `#nullable disable` is inert and is not counted, and its file's nodes are not in the denominator). Each one tells the compiler to trust you about null, suppressing the very safety NRTs provide.
Every external tool invocation behind a deep-scan dimension — the tool, its captured version, the exact command, how many findings it yielded, and a link to the retained raw output. To reproduce any finding: check out the same commit and run the command shown (repo-relative — never an absolute scratch path). The complete raw scanner output is retained verbatim under artifacts/raw/ (indexed in artifacts/raw/index.json); per-invocation exit codes and wall-clock durations are in sidecar.json — kept out of this table so the rendered report stays byte-identical across runs of the same commit.
trivy: not applicable — No Infrastructure-as-Code or container manifests found (Dockerfile, Docker Compose, Terraform, Kubernetes/Helm, CloudFormation, ARM, Bicep, Ansible); nothing to scan.
semgrep: not applicable — No personal data was found crossing a boundary the PII/GDPR ruleset checks — nothing written to a log or console sink, placed in a URL or query string, or persisted to browser storage. That is a clean result for the LEAK surface only: this ruleset detects personal data escaping, it does not inventory the personal data a repository holds, so it is not evidence that this repository has no personal-data surface. The personal-data map (Appendix C) and the C1-C5 compliance cards are what speak to that. semgrep could not parse 12 file(s) — `src/Cai.Web/ApiAccess.cs`, `src/Cai.Web/Noise/FixRateAnchor.cs`, `src/Cai.Web/Noise/MethodVersions.cs`, `src/Cai.Web/Noise/NoiseStandardEndpoints.cs`, `src/Cai.Web/Noise/NoiseStandardHealthCheck.cs`, … (+7 more) — so the PII/GDPR sweep did not cover the unparsed regions of them; rows reported elsewhere in those files are real.
disclosure: not applicable — No vulnerability-disclosure policy file found (SECURITY.md/.markdown/.rst/.txt at root or under .github/.forgejo/.gitea/docs, .well-known/security.txt). A coordinated-disclosure policy may live off-repo, so this is not evidenced rather than failed.
runtime-hardening: not applicable — No Kubernetes/orchestration workloads found in the repository manifests; network egress policy is a cluster-native control that may live at the platform/firewall layer, so there is nothing to assess here.
runtime-hardening: not applicable — No Kubernetes/orchestration workloads found in the repository manifests; seccomp/AppArmor/SELinux confinement is a workload-level control, so there is nothing to assess here.
runtime-hardening: not applicable — No Kubernetes/orchestration workloads found in the repository manifests; runtime threat-detection and admission-control policy are cluster-level controls, so there is nothing to assess here.
0
—
Run 01a0a511-e654-7126-949a-095b25fc2c9c · every finding is also locatable in findings.md, and the complete scoring record (with exit codes + durations) in sidecar.json.
Appendix C — Personal-data map
Every field, property and record parameter whose name is conventional personal data — 2 field(s) across 1 category, each with an exact repo-relative file:line. This is the data inventory a compliance review starts from — right-to-erasure, retention, minimisation. Detected by NAME, from the C# syntax tree, with a deliberately specific identifier classifier — the same one the C1–C5 compliance cards use to decide whether personal data is present, so CardDefinition or FileName don't trip. Two caveats stated rather than glossed: those cards additionally require corroboration (a persistence/account signal, or two distinct PII categories) that this inventory deliberately does not, so it lists more than they gate on; and D32 Data Compliance shares nothing with it — that dimension is a separate semgrep ruleset for personal data leaking into logs, URLs and browser storage, and a clean D32 result says nothing about this list. Informational — it feeds no score.
Issues: 15 · Warnings: 40 · Recommendations: 18 · Info: 19 — Appendix A · all findings · full markdown report.
Generated by Watchdog — deterministic code-health analysis. 15-09-2026 @ 12:36 UTC.
Downloadable artifacts
Machine-readable and reproducible from this commit + frozen rubric — drop them straight into a contract appendix, a CRA dossier, or a downstream SCA / VEX tool.