The Agent Skills Directory

Command Execution: The skill includes a procedure to execute npm run format to maintain project standards. This is a common and expected operation in software development workflows. The skill instructions specify that the agent should ask for user confirmation before running this command, which is a helpful security practice.\n- Data Ingestion Surface: The skill is designed to read and analyze files within the /docs and packages/ directories to ensure documentation reflects the current codebase. This represents a surface where the agent interacts with repository data.\n
Ingestion points: Files in the /docs directory and the packages/ folder (as noted in SKILL.md).\n
Boundary markers: While the skill provides clear stylistic rules for output, it does not explicitly define markers to separate external content during its internal processing.\n
Capability inventory: The skill can modify repository files using file system tools and execute the project's formatting script.\n
Sanitization: The instructions do not specify a process for sanitizing or filtering the content read from repository files before processing.

docs-writer