The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local validation script node scripts/validate-project.mjs and utilizes system tools such as npm view and gh release list to retrieve versioning data. These are standard operations for repository maintenance.
[EXTERNAL_DOWNLOADS]: Retrieves release information from well-known registries including NPM, PyPI, and GitHub repositories. These operations are limited to metadata retrieval from trusted and authoritative sources.
[DATA_EXFILTRATION]: Reads repository manifests and configuration files to determine the technology stack. This data is used solely for internal documentation purposes and is not transmitted to unauthorized external locations.
[PROMPT_INJECTION]: The skill ingests and processes data from external URLs and project files. The risk of indirect prompt injection is mitigated by explicit requirements for version-matched documentation, authority-based verification, and structured constraints on sub-agent tasking.

guidance