The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external data from git commit messages, creating an attack surface for indirect prompt injection. * Ingestion points: Git commit history and local style files (e.g., CHANGELOG_STYLE.md). * Boundary markers: None; there are no specific instructions to separate untrusted commit data from agent instructions. * Capability inventory: Not explicitly defined, but intended for use with agents capable of reading git history and writing files. * Sanitization: No explicit sanitization or filtering of commit message content is performed by the skill itself.

changelog-generator