The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill functions as a static analysis tool for development workflows.
[COMMAND_EXECUTION]: The skill executes git diff commands to identify changed files and content. These are standard, non-privileged operations limited to the local repository context.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted data from code changes and repository guidelines.
Ingestion points: Reads output from git diff main...HEAD --name-only and git diff main...HEAD in SKILL.md.
Boundary markers: Absent; there are no instructions to ignore embedded commands within the diff content.
Capability inventory: Limited to reading the file system and executing git commands. No network access, file-writing, or dynamic code execution (eval/exec) capabilities are present.
Sanitization: Absent; the skill directly scans the raw text of the diff for pattern matching.

pr-readiness-check