brainstorming
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill employs authoritative language such as 'You MUST use this' and 'mandatory' to prioritize the brainstorming workflow. These are standard instructional reinforcements for agent behavior and do not attempt to bypass safety filters or ignore system constraints.- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill reads project context from existing files, documentation, and recent git commits to inform the design process.
- Boundary markers: No explicit delimiters or markers are specified for separating project content from the brainstorming instructions.
- Capability inventory: Capabilities include writing design documents to the
docs/plans/directory, committing changes via git, and invoking other internal skills for workspace management (git worktrees) and plan generation. - Sanitization: No explicit sanitization or filtering of project data is defined.
- Note: This reflects the intended functional design of a development assistant and does not present an elevated security risk beyond standard tool usage.
Audit Metadata