systematic-debugging-hardened
Pass
Audited by Gen Agent Trust Hub on Apr 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides templates for shell commands to diagnose multi-component systems, including macOS keychain inspection and code signing identity verification.
- [PROMPT_INJECTION]: The skill employs imperative language and strict rules to enforce its debugging methodology, which uses patterns similar to behavioral overrides.
- [DATA_EXFILTRATION]: The instructions involve accessing environment variables and keychain metadata for troubleshooting. The skill mitigates exposure risk by directing the agent to use safe presence-checking patterns rather than logging raw secret values.
- [PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data such as error logs and stack traces, presenting a surface for indirect instructions.
- Ingestion points: Error messages, stack traces, and code diffs processed in SKILL.md.
- Boundary markers: None identified.
- Capability inventory: Execution of shell diagnostic commands.
- Sanitization: Specific guidelines for secret masking in diagnostic output are provided in the guardrails section.
Audit Metadata