The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes git and gh CLI commands to extract repository information and manage pull requests. Shell commands in SKILL.md (Step 5) use quoted heredocs (<< 'EOF') to safely handle user-generated content, preventing command injection.
[DATA_EXFILTRATION]: Source code diffs and commit history are retrieved and transmitted to GitHub via the gh CLI. This is the intended primary purpose of the skill and targets a well-known, legitimate service.
[EXTERNAL_DOWNLOADS]: The README provides installation instructions using npx for a scoped package (@opendirectory.dev/skills) and suggests downloading the skill directory via download-directory.github.io.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted data from git logs and diffs.
Ingestion points: Git diff output, commit logs, and existing PR metadata are read into the agent context in SKILL.md (Step 2).
Boundary markers: Absent; the skill does not use specific delimiters to separate code content from instructions.
Capability inventory: The skill possesses write capabilities through gh pr create and gh pr edit (Step 5).
Sanitization: No explicit sanitization or filtering of the git data is performed before processing.
Mitigation: The potential risk is mitigated because the skill is designed to present the generated description to the user for review and confirmation before execution.

pr-description-writer