The Agent Skills Directory

[SAFE]: The skill implements a structured design workflow. It uses standard file operations to save design reports to a .report directory and utilizes git for committing documentation, which are expected behaviors for its stated purpose.
[INDIRECT_PROMPT_INJECTION]: The skill explores existing project files and documentation to gain context for brainstorming. While this provides an entry point for external data, the skill is constrained to conversational design and documentation tasks.
Ingestion points: Reads local project files, documentation, and commit history during the exploration phase (SKILL.md).
Boundary markers: None explicitly defined for ingested content.
Capability inventory: Limited to writing markdown reports to specific directories and executing git commits.
Sanitization: No specific sanitization methods for ingested data are mentioned.

brainstorming