The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted data from external sources such as Git commit logs, Pull Request descriptions, and ticket exports (e.g., Jira, Linear). This content could contain malicious instructions intended to manipulate the agent's summarization or influence subsequent file-writing operations.
Ingestion points: SKILL.md (Step 3: Gather evidence from merged PRs, issues, commits, and ticket exports).
Boundary markers: The instructions lack specific delimiters or warnings to ignore embedded instructions within the processed data, though Step 8 includes a quality checklist to verify claims.
Capability inventory: The skill possesses file-system access capabilities (Read, Write, Edit, Glob, Grep) as defined in the allowed-tools metadata in SKILL.md.
Sanitization: There are no explicit instructions for sanitizing or escaping the content retrieved from external sources before it is used to generate output.

changelog-maintenance