workon
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill description specifies that it loads content from external files (AGENTS.md) and tool outputs (git status, td issues) into the agent's working context.
- Ingestion points: Data is pulled from AGENTS.md and the output of shell tools like git.
- Boundary markers: The skill lacks explicit delimiters to prevent the agent from executing instructions found in the loaded data.
- Capability inventory: The skill invokes the workon tool to change the agent's environment.
- Sanitization: No sanitization or verification of the ingested content is performed.
Audit Metadata