aldescription
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill uses directive language ('Mandatory') to override the agent's default 'plan mode' behavior, insisting on the execution of a specific 'description protocol'.
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface: 1. Ingestion points: Analyzes 'completed changes' from PR/MRs. 2. Boundary markers: None provided to separate untrusted PR data from instructions. 3. Capability inventory: No code execution or file system capabilities identified in the skill file. 4. Sanitization: No input validation or instruction-ignoring directives are present to handle malicious content in PR data.
- [NO_CODE]: The skill does not contain any executable code or scripts, relying entirely on natural language instructions.
Audit Metadata