Anti Entropy Governance
Picks delete-vs-retain when retiring old code, and hard-stops before irreversible data deletion
Test report
- Verdict
- Tested · Works
- Score
- Tested
- Jul 21, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 127eba0
⚠ The author changed this skill after we tested it. The verdict below describes the version we ran on Jul 21, 2026; it's queued for a re-test.
Fetched the raw SKILL.md (HTTP 200) and confirmed it references no external scripts/files (only sibling skill names), so nothing to break; installed cleanly into a temp HOME. Ran one representative task mixing an internal keyword-fallback retirement with a stale users.legacy_locales column drop. Baseline artifact treated both as one cleanup and offered to run ALTER TABLE users DROP COLUMN legacy_locales, verifying only via "tests green"; the skill-followed artifact classified the column as persistent-state, emitted the Data Destruction Guard, refused to run the DROP as the next action, demanded environment+backup+reader-check+scoped confirmation, and separately proceeded with delete-first on the code plus a real negative/lingering-reference verification plan. Artifacts at scratchpad/baseline_artifact.md and scratchpad/skill_artifact.md. No security smells; the skill is itself anti-destructive.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 5/5
- Output vs. baseline 8/10
- Docs & honesty 5/5
What Anti Entropy Governance does
A governance skill for code-retirement tasks: it classifies each deletion target (internal code vs schema/contract vs persistent live data) and picks delete-first, compat-exception, or confirmation-first accordingly. Triggers when you retire old logic, collapse duplicate owners, remove fallbacks, or touch schema, persistence, or source-of-truth boundaries, and it blocks irreversible data deletion (DROP TABLE/COLUMN, TRUNCATE, bulk deletes) until you give explicit scoped confirmation.
How to install Anti Entropy Governance
git clone --depth 1 https://github.com/GanyuanRan/Aegis.git /tmp/anti-entropy-governance-src
mkdir -p ~/.claude/skills
cp -R /tmp/anti-entropy-governance-src/skills/anti-entropy-governance ~/.claude/skills/anti-entropy-governance
# Single-file skill (SKILL.md only), no scripts/deps. MIT licensed.
# Part of the larger "aegis" plugin but the skill dir is self-contained and works standalone.
# Designed to be composed by other Aegis skills (brainstorming/writing-plans/systematic-debugging/verification-before-completion); also works on its own.
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Anti Entropy Governance
-
/anti-entropy-governancePicks delete-vs-retain when retiring old code, and hard-stops before irreversible data deletion
It also activates on plain-language prompts like these:
-
Should I delete this deprecated fallback path -
Review whether it's safe to remove this old logic -
Check if this schema change breaks source-of-truth boundaries
Frequently asked questions
- Is the Anti Entropy Governance skill free?
- Yes. The skill itself is free from GanyuanRan/Aegis. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Anti Entropy Governance work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 21, 2026. Verdict: Tested · Works. Fetched the raw SKILL.md (HTTP 200) and confirmed it references no external scripts/files (only sibling skill names), so nothing to break; installed cleanly into a temp HOME. Ran one representative task mixing an internal keyword-fallback retirement with a stale users.legacy_locales column drop. Baseline artifact treated both as one cleanup and offered to run ALTER TABLE users DROP COLUMN legacy_locales, verifying only via "tests green"; the skill-followed artifact classified the column as persistent-state, emitted the Data Destruction Guard, refused to run the DROP as the next action, demanded environment+backup+reader-check+scoped confirmation, and separately proceeded with delete-first on the code plus a real negative/lingering-reference verification plan. Artifacts at scratchpad/baseline_artifact.md and scratchpad/skill_artifact.md. No security smells; the skill is itself anti-destructive.
- What is the Anti Entropy Governance SkillProof Score?
- 9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
- How do I install Anti Entropy Governance?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Anti Entropy Governance with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.