doc-sys-validator
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script located at
ai_dev_ssd_flow/06_SYS/scripts/validate_sys.pyto perform document validation tasks. - [PROMPT_INJECTION]: The skill processes untrusted external data in the form of SYS documents, which constitutes an indirect prompt injection attack surface.
- Ingestion points: Files located within the
docs/06_SYS/directory are parsed and validated. - Boundary markers: The instructions do not define clear delimiters or markers to separate the content of the SYS documents from the agent's instructions.
- Capability inventory: The skill has the capability to execute shell commands (running the Python validator script).
- Sanitization: There is no mention of sanitization or filtering logic to handle malicious prompt fragments that might be embedded within the documents being validated.
Audit Metadata