The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external sources such as pull request descriptions, commit messages, and issue tickets. This creates a surface for indirect prompt injection where an attacker could embed malicious instructions within these records to influence the generated release notes.\n- Ingestion points: Pull request bodies (via gh pr list), commit messages (via git log), and ticket descriptions (Linear/Jira/GitHub).\n- Boundary markers: The skill lacks explicit instructions to disregard or isolate natural language instructions found within the processed data.\n- Capability inventory: The skill utilizes read-only commands (git, gh) and does not possess capabilities for file modification, arbitrary code execution, or non-whitelisted network exfiltration.\n- Sanitization: No sanitization or filtering of the source text is performed before processing.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using standard development tools to retrieve change history and metadata.\n- Evidence: The skill uses git log and various GitHub CLI (gh) commands to fetch data for analysis.

release-notes-writer