github-issue-creator
Pass
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: SAFENO_CODEDATA_EXFILTRATION
Full Analysis
- [NO_CODE]: The skill consists exclusively of natural language instructions and markdown templates. No scripts or executables are included.
- [SAFE]: The skill includes explicit instructions to redact sensitive information using placeholders like [PROJECT_NAME] and [USER_ID], preventing accidental data exposure.
- [DATA_EXFILTRATION]: The skill instructs the agent to write files to a local directory (/issues/) but does not initiate any network connections or external data transfers.
- [SAFE]: Regarding indirect prompt injection, the skill ingests unstructured user data (Ingestion points: error logs, voice notes). While it lacks explicit boundary markers, its capabilities are restricted to writing markdown files (Capability inventory: file-write to /issues/). It provides sanitization guidance for sensitive data (Sanitization: use of placeholders).
Audit Metadata