verified-spec-code-mapper
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is composed of documentation and guidelines for a proof-auditing workflow. No malicious patterns, obfuscation, or unauthorized commands were detected.
- [PROMPT_INJECTION]: The skill includes instructions to ingest and analyze user-provided source code (Coq, Dafny, Isabelle). Ingestion points: User codebase files; Boundary markers: None; Capability inventory: Limited to generating Markdown reports without shell or network tools; Sanitization: None. This surface is evaluated as safe due to the lack of exploitable capabilities within the skill's instruction set.
- [EXTERNAL_DOWNLOADS]: All external URLs referenced in the skill point to official, well-known documentation for formal verification tools (Coq, Dafny, and Isabelle).
Audit Metadata