agentsclimarketplace

Spec audit docs

Skill ainova-systems/intelligence-dev-packs/packs/spec/skills/spec-audit-docs

Intelligence pack for software engineering based on Ainova Systems best practices

Install
npx -y skills add ainova-systems/intelligence-dev-packs --skill spec-audit-docs

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 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.

What its author says it does

Copied from the file, not written here

Audit documentation claims against the code; report drift vs violation

SKILL.md

1.6 KB, as published. Nobody here has run it

Docs Sync Check

Docs that contradict code are worse than none - readers and agents act on them. Read-only by default.

Steps

  1. Scope: the argument; else the primary docs (README, the docs tree, rules, feature docs, spec claims).
  2. Extract verifiable present-tense claims: commands, paths, endpoints, config keys, versions, architecture statements ("X calls Y", "all writes go through Z").
  3. Verify each against the repo: the command exists in the manifest? the path exists? the route matches the definitions? the boundary holds in the imports?
  4. Classify each mismatch: drift (code moved on - fix the doc) / violation (the doc states the rule, the code broke it - fix the code or escalate; never silently rewrite the rule to match the violation) / ambiguous (owner call).
  5. Report ordered by reader impact (broken setup instructions first): doc location, the claim, the actual state with evidence, classification, suggested fix.
  6. Fix doc-side findings on request; code-side fixes go through the normal flow.

Verify

  • Every reported claim has evidence on both sides (doc line + repo state).

Scope / hand-off

  • Post-execution doc updates - spec-execute Phase D; decisions - spec-decision.

CRITICAL

  • Intentions and roadmap lines are not checkable claims - skip them.
  • Generated docs - report drift against the generator's input; regenerate, never hand-edit the output.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.