The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes external skill files (e.g., in config/source/skills/) for quality and redundancy audits.
Ingestion points: The skill reads markdown files from the local config/source/skills/ directory as specified in the audit workflow.
Boundary markers: No explicit boundary markers or isolation instructions are defined for processing the analyzed content within the audit instructions.
Capability inventory: The skill references local command execution via npm run build:skill-manifest and file system reads across the skill collection.
Sanitization: No specific sanitization or filtering of the ingested skill content is mentioned before analysis.
[COMMAND_EXECUTION]: The instructions include running a local build command npm run build:skill-manifest to generate skill manifests for plugin deployment.
[EXTERNAL_DOWNLOADS]: The documentation references established resources and best practices from well-known services such as Anthropic's platform documentation, GitHub, and Trae's official documentation.

skill-authoring