haskell-pro
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions guiding the agent on how to provide Haskell programming assistance. No executable scripts, binary files, or system commands are present within the skill content.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided Haskell tasks and code. While this constitutes a standard attack surface, the instructions emphasize domain modeling and type safety, and direct the agent to ask for clarification if success criteria or safety boundaries are missing.
- Ingestion points: User-provided Haskell goals and requirements (SKILL.md).
- Boundary markers: Absent.
- Capability inventory: No tools, subprocesses, or network operations are utilized.
- Sanitization: Absent.
Audit Metadata