future-grilling
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted markdown data from local files to influence its interviewing behavior.
- Ingestion points: The agent is instructed to look for and read brief files at 'docs/point-of-view/' and '~/.claude/point-of-view/' based on a user-provided slug.
- Boundary markers: No boundary markers or 'ignore embedded instructions' warnings are present to delimit the brief content from the agent's core instructions.
- Capability inventory: The skill directs the agent to look up facts using filesystem, code, and documentation tools instead of asking the user, providing a broad read access capability.
- Sanitization: There is no description of sanitization, escaping, or filtering of the external brief content before it is interpolated into the agent's context.
Audit Metadata