The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the crit CLI tool to fetch authoring guides, prepare metadata, and manage the review lifecycle through commands like crit story and crit comment. It also executes the tool in the background to listen for user completion.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external feedback (review comments) and instructs the agent to modify source code accordingly. This introduces an attack surface where a malicious reviewer could attempt to inject instructions into the comment stream to manipulate the agent's file-editing behavior. However, this is the primary intended purpose of the skill and is governed by standard LLM guardrails.

crit-story