use-gunshi-cli
Pass
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill directs the agent to read documentation from node_modules/@gunshi/docs/**.md, which serves as a surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion point: Local documentation files located at node_modules/@gunshi/docs/**.md.
- Boundary markers: Absent. The skill lacks instructions to treat documentation content as data rather than instructions.
- Capability inventory: None. No subprocess calls, network operations, or file-write capabilities are defined within this skill.
- Sanitization: Absent. There is no requirement for the agent to sanitize or validate documentation content before processing.
- [NO_CODE]: The skill consists entirely of markdown instructions and does not include any scripts or executable code.
Audit Metadata