Skip to content
cccskills
FreePublished on GitHub

slop-audit

6-lens diagnostic slop audit — unused imports, dead private fns, duplicate signatures, abandoned TODOs, lint bypass markers, stale refs.

The English summary is being prepared. Showing the original description.

See how to install

Files included (1)

  • SKILL.md3.7 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Slop Prevention Audit

Recurring 6-lens scan of core/ / evals/ / evolve/ / scripts/ to catch the patterns that accumulate during long PR sequences and rot codebase health when nobody is grepping for them.

When to run

  • Per session, before the final PR merges. Catches inline duplicates and dead-helper candidates introduced during a sprint.
  • After a large refactor / removal PR. Surfaces stale references to the removed feature in unrelated files.
  • Before promotion. Use it to find review candidates; its counts do not decide promotion.

The 6 lenses

#LensWhat it surfacesSeverity tier
1unused_importsruff F401 — module imports never referenced.warning
2dead_private_functionsdef _foo(...) with zero in-file callers.warning if >10
3duplicate_signaturesSame function name defined in ≥3 files (lift-to-helper signal).info
4abandoned_todosTODO/FIXME/XXX without owner (name) or YYYY-MM-DD date.warning if >5
5lint_bypass_markers# noqa + # type: ignore count.info
6stale_referencesKnown-removed names that re-appear in source (BudgetGuard, seeds_safe10, etc.).error if >0

Severity is diagnostic. Heuristic absolute counts are not an accepted-debt floor or promotion contract.

Reading the output

uv run python scripts/slop_audit.py

Emits a markdown report with the count table + first 5 samples per lens. Inspect samples for patterns:

  • unused_imports samples in tests/ → leftover after a refactor. Auto-fix: uv run ruff check --fix --select F401.
  • dead_private_functions in agent files → same-file textual candidates. Check cross-module, registry, entrypoint, and compatibility consumers before changing them.
  • duplicate_signatures = "init"/"build"/etc. → expected framework method. Other shared names still require behavior and caller comparison; they do not automatically justify a shared helper.
  • abandoned_todos → add owner (@handle) or remove. The convention is # TODO(@user, 2026-05-18): ....
  • stale_references = N for any N > 0 → fix in the same PR as the feature removal; allowing it to land merges the rot.

Workflow

uv run python scripts/slop_audit.py

The original 2026-05-18 absolute-count snapshot is preserved at docs/reference/2026-05-18-slop-audit-baseline.md as historical evidence. It is not a current pass/fail contract. Ruff, mypy, dependency checks, coverage, and behavior tests remain the promotion gates.

Treat all six lenses as discovery signals. Before deleting code or tests, use the deletion gate in .agents/skills/agent-anti-pattern/references/field-guide.md.

Stale-reference list

scripts/slop_audit.py maintains _STALE_REFS — a tuple of names that should not re-appear in production code after their removal PR. Current list:

  • BudgetGuard (removed PR 1)
  • SUBAGENT_BUDGET_WARNING (removed PR 1)
  • seeds_safe10 (renamed PR 0)
  • FitnessBaseline (removed S9)
  • plugins.benchmark_harness, plugins.crucible, plugins.petri_audit, and plugins.seed_generation (removed in v1.0.24)
  • core.self_improving and geode_product (removed in v1.0.24)

Add a new entry whenever you remove a public surface. The lens ignores docs/ and CHANGELOG.md so historical prose stays documented. Independent docs/package/install probes mark only their intentional retirement expectations with # slop:keep.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Audit agent runtimes and their tests for lifecycle inference, capability overload, context pollution, correlated verification, prompt-only control, false success, dead compatibility paths, test-only seams, and complexity without contribution. Use for whole-repository agent audits, slop or dead-code triage, test contribution reviews, large refactors, tool/context simplification, and cleanup requests that need proof before deleting code or tests.

The English summary is being prepared. Showing the original description.

mangowhoiscloud/geode15Updated Oct 22, 2026

Diagnose agent operational failures through default-policy checks, multi-gap analysis, explicit service ownership, request-context lifetime, and degradation-versus-correctness checks.

The English summary is being prepared. Showing the original description.

mangowhoiscloud/geode15Updated Oct 22, 2026

Align, run, audit, and publish GEODE benchmark comparisons against Agent-World arXiv 2604.18292v1 across MCP-Mark, BFCL V4, and tau2-bench. Use for Agent-World or AgentWorld comparison requests, cross-suite agent benchmark design, Tau2 comparator/profile diagnosis, paired thin-runtime-versus-GEODE experiments, mean-accuracy@8 replication, or benchmark comparability and artifact reviews.

The English summary is being prepared. Showing the original description.

mangowhoiscloud/geode15Updated Oct 22, 2026

Verification checklist to prevent fake success. Detects test deletion/disabling, coverage regression, lint bypass, secret exposure. Triggered by "deception", "fake" ("가짜"), "fake success", "verification" ("검증"), "checklist" ("체크리스트"), "deletion detection" ("삭제 탐지"), "regression" keywords.

The English summary is being prepared. Showing the original description.

mangowhoiscloud/geode15Updated Oct 22, 2026

Implement proven backend architecture patterns including Clean Architecture, Hexagonal Architecture, and Domain-Driven Design. Use when architecting complex backend systems or refactoring existing applications for better maintainability.

The English summary is being prepared. Showing the original description.

mangowhoiscloud/geode15Updated Oct 22, 2026

Auto-search and summarize latest AI/agent papers.

The English summary is being prepared. Showing the original description.

mangowhoiscloud/geode15Updated Oct 22, 2026

See all skills by mangowhoiscloud

Report a problem with this skill