gitlab-automation

Warn

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to add a third-party URL (https://rube.app/mcp) as an MCP server configuration. This introduces a remote dependency that executes automation logic and handles GitLab authentication flows through an external service.
  • [PROMPT_INJECTION]: The skill processes data from external GitLab projects, which provides a surface for indirect prompt injection attacks.
  • Ingestion points: Data retrieved through tools such as GITLAB_LIST_PROJECT_ISSUES, GITLAB_GET_PROJECT_MERGE_REQUESTS, and GITLAB_LIST_REPOSITORY_COMMITS as specified in SKILL.md.
  • Boundary markers: The instructions lack guidance for the agent to use delimiters or explicit "ignore embedded instructions" warnings when handling content from issues or merge requests.
  • Capability inventory: The skill allows the agent to create and update projects, issues, branches, and merge requests based on input data (SKILL.md).
  • Sanitization: There are no specified procedures for sanitizing, validating, or filtering the data ingested from GitLab before it is used in subsequent agent operations.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 31, 2026, 04:34 PM
Security Audit — agent-trust-hub — gitlab-automation