メニュー
Use when you need QA-native agent security (taint tracking, capability tokens, deterministic obstructions) to gate tool execution or assess prompt-injection risk while working in this repo.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | qa-agent-security |
| description | Use when you need QA-native agent security (taint tracking, capability tokens, deterministic obstructions) to gate tool execution or assess prompt-injection risk while working in this repo. |
| metadata | {"short-description":"QA agent security guardrails"} |
This skill points you at the repo’s existing QA-native security kernel and how to use it.
Policy kernel self-test (JSON output):
python qa_agent_security/qa_agent_security.py --validateTool runner self-test:
python -m qa_agent_security.tool_runnerFull test suite:
python -m pytest qa_agent_security/tests/ -q --override-ini="testpaths=qa_agent_security/tests" --override-ini="python_files=test_*.py"qa_agent_security/README.mdqa_agent_security/qa_agent_security.pyqa_agent_security/tool_runner.pyShow certificate ecosystem status — family count, recent additions, any gaps
Open Brain status — recent thoughts, stats, and any items needing action
Run the full QA validation suite — axiom linter + meta-validator
Run the weekly review pipeline — activity log + scorecard + review draft
Use when chat/forensics mention scripts or paths that are missing from the current tree, to scan repo zip snapshots and locate where those missing files likely live (outputs a report under `_forensics/`).
Use when working in `signal_experiments/` to build/query repo-wide forensic indices (forensics report, SQLite FTS, QA-certified datastore+views) and to keep `Documents/RESULTS_REGISTRY.md` evidence-linked and searchable.