The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads and installs various ecosystem utility tools at runtime from official package registries and well-known domains, including cargo-semver-checks via crates.io, @arethetypeswrong/cli and @microsoft/api-extractor via npm, gorelease via golang.org, and griffe via PyPI.
[COMMAND_EXECUTION]: The skill executes multiple system commands to manage the versioning workflow, including git for diffing and tagging, and package managers like npm, cargo, poetry, and hatch for manifest updates. It also references vendor-specific scripts like publish-skill and release.py for managing skill releases.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain malicious instructions. * Ingestion points: It reads raw git diff output and the results from external breaking-change detectors to make versioning decisions. * Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish untrusted tool output from internal logic. * Capability inventory: The skill has the ability to modify project files and execute shell commands through various package managers and release scripts. * Sanitization: There is no evidence of sanitizing or filtering the content retrieved from git diffs or tool outputs before it is analyzed by the agent.

semver