agentleFS
Sign inStart free
r-irber-irbe/proof-skills/skills/lean-deep-audit/SKILL.md

lean-deep-audit

Skill3 starsChanged 24 days agoApache-2.0View on GitHub

USE FOR: running a systematic whole-corpus quality audit over a large Lean formalization project: trust-base lockdown. DO NOT USE FOR: reviewing a single theorem's proof (@lean-proof-review). TRIGGERS: deep audit, corpus audit, whole-corpus review.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    P1 - Trust-base lockdown: Every declaration's transitive axiom closure restricted to the declared allowlist, enforced in the normal build.

    line 21
  2. Distinctive:

    If a census script times out or drifts after upstream edits, re-run the precompiled two-pass scanner on a clean tree before quoting line …

    line 34
  3. Distinctive:

    If a module family is guest or template material, register the finding in the remediation queue and escalate disposition to the owning lane …

    line 38

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

The Deep-Audit Program (whole-corpus, seven lenses)

Operational dispatch contract for @lean-deep-audit.

Routing

  • USE FOR: tasks routed to @lean-deep-audit.
  • DO NOT USE FOR: out-of-scope tasks (delegate via @lean-gateway).

Workflow

  1. P1 - Trust-base lockdown. Every declaration's transitive axiom closure restricted to the declared allowlist, enforced in the normal build. EXECUTOR: the lean-axiom-closure skill (vendored metaprogram
    • FinalCheck consumer). Expected first catches: sorryAx behind wrappers/aliases, native_decide (Lean.ofReduceBool), custom axioms.
  2. P2 - Literature faithfulness, tower by tower. For each foundational module: read every core definition; compare with the cited formula; classify faithful-formula / scoped-specialization / free-record-proxy. The proxy signature: a structure with free fields + arithmetic defs over those fields + projection-style theorems, while the module prose narrates mathematics the code

Recovery & STOP

  • If a census script times out or drifts after upstream edits, re-run the precompiled two-pass scanner on a clean tree before quoting line numbers or counts.
  • If a module family is guest or template material, register the finding in the remediation queue and escalate disposition to the owning lane rather than deleting modules unilaterally.

Handoffs

  • Predecessors: @lean-gateway (top-level routing), @lean-quality-engine (project QA orchestration).
  • Successors: @lean-axiom-closure (P1 executor), @lean-tautology-triage (P3 candidate triage), @lean-proof-review (per-module review).

Detailed reference

Full encyclopedic reference, decision trees, and lookup tables live in references/lean-deep-audit-handbook.md .

See also

  • ../../references/lean-deep-audit-handbook.md -- Complete lean-deep-audit Handbook
  • ../lean-gateway/SKILL.md -- Top-level skill router

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free