The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from git logs and diffs, creating an indirect prompt injection surface. 1. Ingestion points: The skill runs git log, git diff, and git show to collect analysis context (SKILL.md). 2. Boundary markers: No explicit delimiters or boundary markers are instructed for use when the agent reads diff content. 3. Capability inventory: The skill utilizes git CLI commands and selective file reading (SKILL.md). 4. Sanitization: No sanitization or filtering is performed on the git output before the agent analyzes it.
[COMMAND_EXECUTION]: The skill executes shell commands using the git CLI to inspect repository history and changes. While these are necessary for the skill's functionality, they involve direct interaction with the system's command line environment.

lesson-learned