coding
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process where the agent reads and analyzes external codebase files, module structures, and third-party documentation.
- Ingestion points: Codebase files, module structure, and third-party source/documentation (SKILL.md).
- Boundary markers: The skill does not implement delimiters or specific instructions to ignore embedded instructions in analyzed code.
- Capability inventory: Modifying code, performing git commits, and executing verification tests (SKILL.md, references/debugging.md).
- Sanitization: No sanitization or filtering procedures are defined for the ingested data.
- [EXTERNAL_DOWNLOADS]: The skill documentation includes metadata referencing information sources from established technology organizations and academic repositories (.dev/reference-inventory.json); these are treated as safe information references.
- [SAFE]: The skill content is purely instructional and does not include any executable scripts, automated system commands, or mechanisms for persistence.
Audit Metadata