The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/analyze_changes.py utility executes git via subprocess.run to retrieve repository metadata and changes. The implementation correctly avoids shell=True, mitigating shell injection risks, and execution is scoped to the local repository.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the analysis of untrusted git diff content.
Ingestion points: The scripts/analyze_changes.py script ingests content from git diff output in get_diff_content and detect_high_priority_changes.
Boundary markers: Findings from the diff are printed directly to the agent's context without delimiters or instructions to ignore embedded commands.
Capability inventory: The agent using this skill typically has permissions to modify the file system and execute tools based on the instructions in SKILL.md.
Sanitization: The script extracts and displays raw line segments from the diff in detect_high_priority_changes without sanitization, allowing embedded instructions to reach the agent.

documentation-reviewer