eng-ai
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill definition does not contain any detected security threats, malicious instructions, or obfuscated content.
- [COMMAND_EXECUTION]: The workflow documentation specifies the use of benign shell commands (cat, ls, grep) for analyzing local project files within designated directories like ai/memory-bank/ and data/.
- [PROMPT_INJECTION]: The skill has an indirect injection surface as it ingests untrusted data and has exploitable capabilities.
- Ingestion points: Reads ai/memory-bank/requirements.md, ai/memory-bank/data-sources.md, and files in data/ (SKILL.md).
- Boundary markers: None identified.
- Capability inventory: Uses bash, git, and file operations (SKILL.md).
- Sanitization: None identified.
Audit Metadata