Goal Audit — Security skill for Claude Code
Attack the goal's own acceptance criteria — ledger integrity, negative control, and mutation probe (does not consume an iteration).
How to install Goal Audit
Installs to ~/.claude/skills/nova-violet-role-rot-dtd-goal-goal-audit/SKILL.md
mkdir -p ~/.claude/skills/nova-violet-role-rot-dtd-goal-goal-audit && curl -fsSL https://raw.githubusercontent.com/Nova-Violet-Role/RoT-DTD-GOAL/HEAD/commands/goal-audit.md -o ~/.claude/skills/nova-violet-role-rot-dtd-goal-goal-audit/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Goal Audit does
description: Attack the goal's own acceptance criteria — ledger integrity, negative control, and mutation probe (does not consume an iteration)
G="${CLAUDE_PLUGIN_ROOT}/scripts/goal.sh"
bash "$G" audit ; echo "audit exit=$?"
bash "$G" redteam ; echo "redteam exit=$?"
bash "$G" mutate ; echo "mutate exit=$?"
Three different questions, in increasing strength:
- audit — do the criteria still hash to what was sealed at activation? Drift means someone edited a che
Alternatives in Security
- Clawsec — Security skill suite with drift detection, automated audits, and skill integrity verification 798 ★
- Text Integrity Audit.Skill — text-integrity-audit.skill.md 265 ★
- Goal Prompt Builder — Build audit-friendly /goal prompts for OpenAI Codex 225 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Review Test Quality
Audit test suite quality: coverage gaps, weak assertions, missing negative cases, over-mocking, and tests that
Aegis Activity Logger
Append-only JSONL audit log of every Edit/Write/Bash mutation, one file per UTC day. Provides historical repla
Agent Audit
Audit a clarc agent file for quality across 8 dimensions — instruction clarity, model appropriateness, tool co
Audit Spec
Audit dark-factory spec file against preflight checklist and quality criteria
Spec Check
Audit GWT acceptance test specs for implementation leakage. Optionally provide a specific file path.
Audit Cron Arc
Meta-analyze the effectiveness of a /loop arc — per-iteration metrics, LOC delta trend, saturation detection,
Related Agents
Testwriter
Writes tests that map 1:1 to spec.md acceptance criteria. Runs tests and mutation analysis. Produces tests.md.
Ae Test
Agent Engineer test agent. Writes tests from the acceptance criteria rather than from the implementation, muta
Goal Redteam
Attacks an acceptance criterion. Invoke when a criterion looks too easy, survived the empty-directory control,