feature-forge
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input during discovery and interview sessions to produce technical specifications and checklists, creating an injection surface.
- Ingestion points: User responses gathered via the AskUserQuestions tool and chat-based elicitation.
- Boundary markers: The skill does not implement explicit delimiters to distinguish user-provided content from agent instructions in the output documents.
- Capability inventory: The agent can write markdown files to the specs directory and invoke other agents using the platform's multi-agent delegation features.
- Sanitization: User input is interpolated into final specification templates without visible sanitization or validation logic.
Audit Metadata