The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes content from project-level instruction files to determine which design decisions have already been made and which are still open.
Ingestion points: The skill reads from user-supplied instructions, project agent instruction files (specifically CLAUDE.md and AGENTS.md), and files within decision record or Architecture Decision Record (ADR) directories.
Boundary markers: There are no specific delimiters or instructions to ignore embedded commands mentioned for the reading of these external files.
Capability inventory: The skill is designed to write new markdown files to the local file system (specifically in docs/decisions/ or by modifying CLAUDE.md/AGENTS.md) and hands off control to other domain-specific skills for design and coding tasks.
Sanitization: The instructions do not specify any sanitization, filtering, or validation of the content read from the project instruction files before it is processed by the agent.

godot-grill