to-spec
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes the current conversation context as its primary input to generate specifications. This creates a surface for indirect prompt injection, where malicious instructions hidden in the chat history could be executed by the agent during synthesis or included in the final published specification.
- Ingestion points: The skill explicitly uses the "current conversation context" as defined in
SKILL.md. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the conversation history as untrusted data or to ignore embedded commands.
- Capability inventory: The skill requires the agent to explore the repository (file system read access) and publish to an external issue tracker (network write access).
- Sanitization: No sanitization, filtering, or validation steps are defined for the input conversation data before it is processed or published.
Audit Metadata