Auditing compliance and provenance
Skill brandondees/code-quality-atlas/skills/auditing-compliance-and-provenance
Audits compliance, licensing, and provenance: dependency licenses compatible with the distribution model, copyleft contamination, missing SPDX headers and attribution, code of unclear provenance, PII data flows without minimization or retention limits, consent gating for telemetry, and SBOM currency. Detects and escalates to humans rather than deciding legal questions. A repo-wide / scheduled audit. Use when auditing licenses, PII handling, data retention, or provenance.From its SKILL.md
npx -y skills add brandondees/code-quality-atlas --skill auditing-compliance-and-provenanceAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 9 stars9 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
6.4 KB, as published. Nobody here has run it
auditing-compliance-and-provenance
Any licensing, PII, or provenance exposure? Detect and escalate to humans — never decide legal questions.
When to use
Audits compliance, licensing, and provenance: dependency licenses compatible with the distribution model, copyleft contamination, missing SPDX headers and attribution, code of unclear provenance, PII data flows without minimization or retention limits, consent gating for telemetry, and SBOM currency. Detects and escalates to humans rather than deciding legal questions. A repo-wide / scheduled audit. Use when auditing licenses, PII handling, data retention, or provenance.
Shape: repo. Run against the whole repository (scheduled or on demand), not a single diff.
Reviewer discipline
Report only real problems. If this lens applies and what you reviewed holds up — the code, the design, or the repository's current state — reply "No findings" and stop. If what you were given is outside this lens's scope entirely, say so in one line instead, starting with the words "Not applicable:" followed by what's missing — never the healthy-scan sentence, which means a check ran and found nothing, not that nothing here applied. Either way, do not invent issues. This guards against false positives on correct code; still report every genuine issue you do find, with its full detail.
Defects are the default; improvements are opt-in. By default this lens is defect-only: do not suggest changes to code that is already correct. When the team has opted up into improvement suggestions, a finding on already-correct code is admissible only as nit-severity, route: implementer (the author applies, defers, or ignores), and must clear the non-configurable anti-churn floor: it must genuinely improve — never offer a merely equivalent alternative — and must converge (once a dimension is as good as you can confidently make it, stop; never oscillate A→B then B→A, never re-order to an equivalent state). Defects keep the strict bar above regardless of this setting.
Team preferences. If the reviewed repo has .code-quality-atlas/preferences.md, apply it before reporting: a repo's .code-quality-atlas/preferences.md may set/tune this lens's thresholds or selection, and — being preference-tier — may suppress one of its findings outright (it never surfaces). Its improvement-valence directive is also what decides whether the "opted up" improvement-suggestion behavior above is active for this review. Absent the file, apply this lens's defaults exactly as written above.
Top checks
The head of the full checklist — enough for a first pass without opening any reference file:
- New dependency license check: does an added dependency (and its transitive tree) carry a license compatible with the project's distribution model? Block/strongly-flag GPL/AGPL pulled into a permissive or proprietary product.
- Copyleft contamination / linkage: does the change link or combine with copyleft code in a way that triggers obligations (esp. AGPL over a network service)? Static vs. dynamic linking and "mere aggregation" matter — flag for legal if unsure.
- License/attribution preservation: are upstream license texts, copyright notices, and NOTICE files retained when vendoring/copying code? Removed attribution = violation.
- Provenance of copied code: does this diff include code pasted from Stack Overflow, a blog, another repo, or AI generation without attribution/license clarity? Treat as untrusted: verify license, run secret/IP scan, label it.
- Per-file license header: do new source files have an
SPDX-License-Identifier+ copyright (REUSE)? Missing header = provenance gap. - PII data-flow: does new code collect, store, log, or transmit personal data? If so — lawful basis/consent present, minimized to what's needed, not logged in plaintext, and not sent to a third party/region without basis (cross-links #14, #16, #25).
- Retention & deletion: is there a retention limit and an erasure path for new personal data (right-to-be-forgotten), or does it accumulate indefinitely?
- Data residency / cross-border: does the change move PII to a new region, vendor, or third-party model (incl. LLM APIs) — is that transfer permitted and documented? (cross-links #25 PII-to-model.)
Mechanizing these checks
Where a finding here is one a tool can catch deterministically, surface that as an advisory route: implementer note next to the finding: the hand review caught it this time, and wiring the matching tool from reference/tool-rules.md into CI catches it automatically from then on. This is a suggestion to mechanize, not a defect — it never blocks a verdict, and it falls away on a repo that already runs the tool.
Process notes. If this lens misfired on this change — flagged correct code, missed an obvious issue squarely in its own scope, or its checklist didn't fit the change shape — say so in one line under synthesizing-review-findings's Process notes appendix; that is not a defect finding. Say nothing if the lens worked as intended — never invent a process note to fill the section.
Going deeper
- reference/heuristics.md — the full checklist; open it when the change sits squarely in this lens's domain.
- examples.md — concrete good/bad findings, and the output format to match.
- reference/tool-rules.md — static-analysis rules covering the mechanical subset; for wiring up linters, not needed for the judgment review itself.
- reference/sources.md — the research behind each check; for provenance, not needed during a review.