The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize external developer data such as commit messages, pull requests, and changelogs, creating a surface for potential indirect injection attacks.
Ingestion points: Workflow steps in SKILL.md and the input list in references/release-notes-checklist.md (commit range, PRs, issues).
Boundary markers: None are explicitly defined to isolate untrusted external data within the prompt context.
Capability inventory: The skill is restricted to text generation; there are no file system write operations, network requests, or subprocess executions defined.
Sanitization: No specific filtering or escaping mechanisms are implemented for the external input data.
[SAFE]: No security issues, malicious patterns, obfuscation, or high-risk operations were detected. The skill's primary function is text generation based on user-provided documentation.

release-notes