The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes the stagereview and git command-line tools to process local repository data and present the final review output.
[EXTERNAL_DOWNLOADS]: User instructions include installing the stagereview package from the NPM registry, which is a resource associated with the skill's author.
[PROMPT_INJECTION]: The skill processes external content from git hunks and commit messages, which presents an indirect prompt injection surface.
Ingestion points: Git diff hunks and commit messages are read from a temporary file generated during the preparation step.
Boundary markers: No specific delimiters or warnings are used to prevent the agent from following instructions that might be embedded in the code diffs or commit messages.
Capability inventory: The agent has the capability to execute shell commands and write to temporary files as part of its core functionality.
Sanitization: The input data is analyzed directly without explicit sanitization or filtering of potential injection patterns.

stage-chapters