The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires users to add https://rube.app/mcp as an MCP server. This domain is not recognized as a trusted organization or well-known service, and adding external MCP endpoints grants capabilities to the agent that are managed by a third party.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval of user-generated content from GitLab, which could contain malicious instructions designed to influence the AI agent's behavior.
Ingestion points: Tools such as GITLAB_LIST_PROJECT_ISSUES, GITLAB_GET_PROJECT_MERGE_REQUESTS, and GITLAB_LIST_REPOSITORY_COMMITS retrieve untrusted data from project issues, merge request descriptions, and commit messages.
Boundary markers: The instructions do not define boundary markers or include specific directives for the agent to ignore instructions embedded within the retrieved GitLab data.
Capability inventory: The skill possesses significant capabilities, including the ability to create projects, modify issues, manage branches, and list project members.
Sanitization: There is no indication that the skill validates or sanitizes the content fetched from GitLab before it is processed by the agent.

gitlab-automation