requirements-analyst
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION] (SAFE): No direct injection or bypass patterns found in the skill instructions.
- [DATA_EXFILTRATION] (SAFE): No access to sensitive files or network endpoints detected.
- [COMMAND_EXECUTION] (SAFE): The skill does not invoke shell commands or scripts.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill ingests untrusted data from external discovery documents. Evidence: 1. Ingestion points: Intake and Brief documents (via file paths). 2. Boundary markers: Absent. 3. Capability inventory: Limited to markdown text generation. 4. Sanitization: Not present. The risk is assessed as LOW because the output is intended for human review and the skill lacks side-effect capabilities.
- [REMOTE_CODE_EXECUTION] (SAFE): No remote scripts or external package installations are referenced.
- [OBFUSCATION] (SAFE): The instructions are clear and human-readable with no encoded content.
- [PRIVILEGE_ESCALATION] (SAFE): No administrative commands or permission modifications are present.
- [PERSISTENCE] (SAFE): No mechanisms for establishing persistence on the system were found.
- [METADATA_POISONING] (SAFE): Metadata accurately describes the skill's purpose without deceptive instructions.
- [DYNAMIC_EXECUTION] (SAFE): The skill does not involve runtime compilation or unsafe deserialization.
Audit Metadata