The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes potentially untrusted data from repository files (e.g., .codestable/attention.md, feature-design documents) and user input to drive triage logic, creating an indirect prompt injection surface.
Ingestion points: SKILL.md instructions to scan and read files within the .codestable/ directory and existing brainstorm records.
Boundary markers: None identified in the provided content to separate ingested data from agent instructions.
Capability inventory: The skill can write files to the local filesystem (.codestable/brainstorms/, .codestable/features/) and orchestrate other high-capability skills like cs-feat-design and cs-roadmap.
Sanitization: No explicit sanitization or validation of the content read from files is mentioned.
[COMMAND_EXECUTION]: The skill explicitly permits the agent to create and run small demo or 'spike' scripts (e.g., Python or TypeScript) to verify facts or API behaviors.
Evidence: The 'Small demo / spike' section in SKILL.md directs the agent to create and run files like spike.py or try-{topic}.ts to resolve technical uncertainties during discussions.

cs-brainstorm