delete-agent
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFENO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill processes output from the LiteLLM listing endpoint, which represents a surface for indirect prompt injection. * Ingestion points: The agent list returned by the curl command in SKILL.md. * Boundary markers: No specific delimiters or ignore-instructions are used for the API output. * Capability inventory: The skill uses Bash with curl for network interaction in SKILL.md. * Sanitization: No explicit sanitization or validation of the API response is documented.
- [COMMAND_EXECUTION]: The skill constructs a deletion command by interpolating a user-provided agent ID directly into the request URL in SKILL.md, which could allow for path traversal or API injection.
- [CREDENTIALS_UNSAFE]: The skill correctly uses an environment variable for the LITELLM_API_KEY, avoiding hardcoded secrets.
- [NO_CODE]: The skill is composed only of instructions in SKILL.md and does not include separate script or binary files.
Audit Metadata