explain-deeply
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill includes instructions to override the agent's default response formatting (preferring paragraphs and structured layers over terse fragments) during the explanation phase. This is a stylistic directive to enhance clarity and does not target safety guardrails or core system instructions.\n- [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to ground its explanations in repository files and external info, creating a surface for indirect prompt injection. However, the skill's lack of high-risk capabilities mitigates this threat.\n
- Ingestion points: Local repository files and external documentation fetched via tools.\n
- Boundary markers: No explicit delimiters are specified to separate external content from the prompt context.\n
- Capability inventory: The skill is limited to generating text explanations and ASCII diagrams; it does not perform shell execution, file writing, or credential access.\n
- Sanitization: No sanitization of ingested content is defined in the instructions.
Audit Metadata