The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of git commit logs to generate changelog entries and release notes. This content is then used to populate release metadata and documentation.
Ingestion points: Git commit history and logs are read via npm run changelog:audit as described in SKILL.md.
Boundary markers: None identified; the instructions do not provide specific delimiters or ignore-directives to prevent the agent from obeying instructions potentially embedded in commit messages.
Capability inventory: The skill has the ability to perform high-privilege actions such as git push to the main branch and gh release create.
Sanitization: No sanitization or validation logic is defined for the commit data before it is processed by the agent and incorporated into commands.
[COMMAND_EXECUTION]: The skill uses variables derived from user input and external data to construct shell commands. Variables like <version> and the generated <changelog section> are passed directly to npm, git, and gh tools.

release