The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitLab merge request descriptions and code diffs which could contain malicious instructions.
Ingestion points: The skill fetches external data using glab mr view and glab mr diff as described in SKILL.md.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the review workflow.
Capability inventory: The agent can read repository data and write comments using the glab tool.
Sanitization: No explicit sanitization or filtering of the fetched content is specified before the AI analysis step.
[COMMAND_EXECUTION]: The skill utilizes the glab CLI for GitLab interactions.
The workflow includes the execution of glab mr view, glab mr diff, and glab mr note commands.
Security best practices are encouraged in the instructions, advising the agent to properly quote or escape message content to prevent shell-related issues.

gitlab-mr-review