The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Python script (scripts/git_weekly_report.py) to execute local git log commands via subprocess.run. This is the intended primary function for extracting commit history and is implemented safely using argument lists instead of shell strings.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Git commit messages (subjects and bodies) to generate its reports. There is a surface where malicious commit content could attempt to influence the agent's behavior during the categorization and summary generation process.
Ingestion points: Git commit logs are extracted by scripts/git_weekly_report.py and provided to the agent as structured JSON.
Boundary markers: The instructions lack specific delimiters (e.g., XML tags) or boundary instructions to isolate untrusted commit text from the system instructions.
Capability inventory: The skill has the capability to execute Git commands and generate Markdown files based on external data.
Sanitization: No sanitization, character filtering, or length-based sanitization (other than truncation) is performed on the commit content before it enters the agent's context.

git-weekly-report-cn