proj-decide
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONNO_CODE
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes the
skillwiki validatecommand to verify the integrity of the generated Architectural Decision Records. This is a local operation for document linting. - [PROMPT_INJECTION]: The skill processes external inputs such as project context and user-defined decision slugs, creating a surface for potential indirect prompt injection.
- Ingestion points: Project context, user-provided architectural decisions, and path slugs in SKILL.md.
- Boundary markers: None present to distinguish untrusted data from instructions.
- Capability inventory: File system writes to projects/, concepts/, and vault logs; execution of the skillwiki CLI.
- Sanitization: No explicit sanitization or validation of the input strings before use.
Audit Metadata