How do I install the AUDIT-repo-reality skill?

Run npx skillmds@latest add ariffazil/audit-repo-reality in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is AUDIT-repo-reality safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: reads secrets. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with AUDIT-repo-reality?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is AUDIT-repo-reality free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published AUDIT-repo-reality?

ariffazil (@ariffazil) published this skill. Their other Agent Skills are listed on their SkillMD profile.

AUDIT-repo-reality by ariffazil — Agent Skill

Name: AUDIT-repo-reality
Rating: 4.5 (4 reviews)
Author: ariffazil

AUDIT-repo-reality

REPO_REALITY_AUDIT::v1.2 — Sovereign doctrine (F13, 2026-08-14). Audit codebase reality, not AI output. Categories: STUB Tiers T1-T4, ORPHAN, SHIM, REALITY_LEAK, AUTHORITY_LEAK, CLAIM_DRIFT (A comment/B architecture/C metric), FAKE_METRIC. Pipeline: SCAN→CONTEXT→REACHABILITY→RUNTIME_IMPACT→VERDICT. HEURISTIC≠MEASUREMENT semantics. Nuclear Rule: no abstract feature without concrete consumer. USE WHEN: audit repo, find stubs, reality audit, orphan/shim detection, verify implementation claims.

ariffazil Updated Sep 10, 2026

File contents

AUDIT-repo-reality — Repository Reality Auditor

Operating posture: NO TRUST. Regex flags, context judges, reachability determines risk, reality determines verdict. Scanner output = CANDIDATES only.

Reality Chain

EXISTENCE → REACHABILITY → EXECUTION → DEPLOYMENT → OBSERVABILITY → REALITY. Weakest link = chain verdict: PHANTOM / DEAD / ORPHAN / STUB / SHIM / FAKE_METRIC / THEATRE.

Stub Tiers (classify before severity)

T1 DOCUMENTARY (comment only, code works) → LOW/REFUTED · T2 DORMANT (exists, zero refs) → ORPHAN · T3 REACHABLE (called, body empty/logs) → HIGH · T4 CONSTITUTIONAL (feeds seal/receipt/vault/ledger) → CRITICAL (authority simulation, false reality factory).

Signatures

SOVEREIGN_TOKEN_THEATRE: Math.random near token generation + approval/sovereign keywords + format-only validation downstream → CRITICAL auto-candidate. COMMENT_LIE: docstring/comment asserts verification the code does not perform. SILENT_FAILURE: try{}catch{}-swallowed authority/bootstrap operations whose downstream failure is untraceable.

Verdicts + Report

Per-finding: FILE / LINE / SEVERITY / EVIDENCE / WHY_CANDIDATE → agent adds WHY_IT_IS_FAKE or REFUTED (context may acquit). Corruption Score = STUBS×5 + ORPHANS×3 + DEAD×3 + FAKE_METRICS×10 + REALITY_LEAK×10 + TODOS×2 + UNUSED×2; weekly trend, rising = falling health. Insufficient evidence → VERDICT=UNKNOWN. Never invent findings. Reality > diagrams. Running code > docs. Execution trace > claims.

---name: audit-repo-realitydescription: REPO_REALITY_AUDIT::v1.2 — Sovereign doctrine (F13, 2026-08-14). Audit codebase reality, not AI output. Categories: STUB Tiers T1-T4, ORPHAN, SHIM, REALITY_LEAK, AUTHORITY_LEAK, CLAIM_DRIFT (A comment/B architecture/C metric), FAKE_METRIC. Pipeline: SCAN→CONTEXT→REACHABILITY→RUNTIME_IMPACT→VERDICT. HEURISTIC≠MEASUREMENT semantics. Nuclear Rule: no abstract feature without concrete consumer. USE WHEN: audit repo, find stubs, reality audit, orphan/shim detection, verify implementation claims.---# AUDIT-repo-reality — Repository Reality AuditorOperating posture: NO TRUST. Regex flags, context judges, reachability determines risk, reality determines verdict. Scanner output = CANDIDATES only.## Reality ChainEXISTENCE → REACHABILITY → EXECUTION → DEPLOYMENT → OBSERVABILITY → REALITY. Weakest link = chain verdict: PHANTOM / DEAD / ORPHAN / STUB / SHIM / FAKE_METRIC / THEATRE.## Stub Tiers (classify before severity)T1 DOCUMENTARY (comment only, code works) → LOW/REFUTED · T2 DORMANT (exists, zero refs) → ORPHAN · T3 REACHABLE (called, body empty/logs) → HIGH · T4 CONSTITUTIONAL (feeds seal/receipt/vault/ledger) → CRITICAL (authority simulation, false reality factory).## CategoriesREALITY_LEAK: claim about reality without traceable source (health=healthy / W3=0.9 with no derivation). AUTHORITY_LEAK: claim of authorization without verifiable authority chain (regex/format/boolean/Math.random as "approval"). CLAIM_DRIFT: stated reality ≠ implemented reality (A comment / B architecture / C metric). HEURISTIC vs FAKE: labeled prior/rule-based estimate = HEURISTIC (innocent); heuristic presented as measurement = REALITY_LEAK.## SignaturesSOVEREIGN_TOKEN_THEATRE: Math.random near token generation + approval/sovereign keywords + format-only validation downstream → CRITICAL auto-candidate. COMMENT_LIE: docstring/comment asserts verification the code does not perform. SILENT_FAILURE: try{}catch{}-swallowed authority/bootstrap operations whose downstream failure is untraceable.## Verdicts + ReportPer-finding: FILE / LINE / SEVERITY / EVIDENCE / WHY_CANDIDATE → agent adds WHY_IT_IS_FAKE or REFUTED (context may acquit). Corruption Score = STUBS×5 + ORPHANS×3 + DEAD×3 + FAKE_METRICS×10 + REALITY_LEAK×10 + TODOS×2 + UNUSED×2; weekly trend, rising = falling health. Insufficient evidence → VERDICT=UNKNOWN. Never invent findings. Reality > diagrams. Running code > docs. Execution trace > claims.

Frequently asked questions