The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using pnpm api-diff. Specifically, it runs validation checks and generates documentation files using --check and --write-doc flags.
[INDIRECT_PROMPT_INJECTION]: The skill processes external YAML data which could potentially contain malicious instructions if the source files are compromised.
Ingestion points: YAML annotation files located under migration/annotations/ (referenced in annotations.md).
Boundary markers: The instructions include a manual verification step in SKILL.md (step 2) to check suggested replacements against implementation and tests.
Capability inventory: The skill performs subprocess calls via pnpm and writes to the file system at migration/v3-to-v4.md (referenced in generation.md).
Sanitization: No automated sanitization or schema validation is mentioned for the ingested YAML content before the documentation generation tool processes it.

migration-guidance