ciso-advisor
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions mandate reading external data from 'company-context.md' without providing security boundaries or instructions to ignore potential malicious prompts within that data.
- Ingestion points: The skill explicitly instructs the agent to read 'company-context.md' in the 'SKILL.md' file.
- Boundary markers: There are no defined delimiters or 'ignore' instructions for processing the content of 'company-context.md'.
- Capability inventory: The skill includes 'risk_quantifier.py' and 'compliance_tracker.py' which have the capability to write data to local CSV files. It also uses the '[INVOKE]' mechanism to trigger actions from other agent roles.
- Sanitization: The skill does not implement or recommend any sanitization or validation logic for the external context before processing.
Audit Metadata