recoup-platform-capture-lesson
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill performs legitimate documentation management tasks by reading and writing to specific workspace paths such as learnings/ to maintain a record of solved problems.
- [COMMAND_EXECUTION]: The agent is instructed to perform git commits to track changes to the learning files, which is a standard and safe development workflow within a codebase.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes task information for persistent documentation that future agents may read. 1. Ingestion points: Reads user-provided solutions and existing files in the learnings/ directory (SKILL.md). 2. Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the documentation templates. 3. Capability inventory: File system write operations and git commit commands (SKILL.md). 4. Sanitization: No explicit sanitization or validation of the captured lesson content is mentioned.
Audit Metadata