The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx to download and execute the mdxg-redline package from the NPM registry. This involves fetching and running external code at runtime to perform the review tasks.
[COMMAND_EXECUTION]: The skill executes bash scripts (scripts/request-review.sh, scripts/wait-feedback.sh) and direct CLI commands (npx mdxg-redline --clean) to manage the review workflow, launch browser-based review interfaces, and poll for feedback files.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface by reading and acting upon reviewer comments stored in feedback.json files. The agent is instructed to interpret these comments as editing commands for the target markdown documents.
Ingestion points: Reads feedback.json in Step 4 of the review process.
Boundary markers: None; the skill does not use specific delimiters or instructions to prevent the agent from obeying instructions embedded within the feedback comments.
Capability inventory: The agent can read and modify local files (using Read and Edit tools) and execute shell commands (via the Bash tool).
Sanitization: There is no evidence of validation or sanitization of the comment content before the agent interprets it for file modifications.

md-review