The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a local script scripts/collect-release-material.sh to query the repository's git history. This involves executing commands such as git log, git diff, and git shortlog to collect the raw data required for the changelog synthesis.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from the repository's development history.
Ingestion points: Commit messages, pull request titles/bodies, and issue descriptions extracted via collect-release-material.sh or manually pasted by the user.
Boundary markers: The instructions do not define explicit delimiters to separate system instructions from the processed data in the prompt context.
Capability inventory: The skill has the capability to execute shell scripts and generate formatted documentation.
Sanitization: While the AI is directed to verify every entry against a code diff and provide a traceability table, there are no automated sanitization filters for embedded instructions within the ingested text.

changelog-writing