The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the local git CLI (via child_process.spawnSync) to collect commit history, remote URLs, and configuration settings from repositories on the local filesystem.
[DATA_EXPOSURE]: Scans local directories to identify Git repositories and reads commit data (messages, dates, hashes) to produce report files. This behavior is documented and central to the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it processes Git commit messages—which could contain malicious instructions—by passing them to an AI agent for summarization.
Ingestion points: Git commit messages retrieved via git log in scripts/weekly.js.
Boundary markers: Absent. The prompt template in resources/prompt.txt does not use explicit delimiters or instructions to ignore embedded commands within the commit data.
Capability inventory: The skill executes git commands and writes files (.json, .docx) to the local filesystem.
Sanitization: No sanitization is performed on the commit messages before they are processed by the agent.

report