gitlab-automation

Pass

Audited by Gen Agent Trust Hub on Jul 18, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and process untrusted data from GitLab repositories.
  • Ingestion points: External data enters the agent context via tools such as GITLAB_LIST_PROJECT_ISSUES, GITLAB_GET_PROJECT_MERGE_REQUESTS, and GITLAB_LIST_REPOSITORY_COMMITS.
  • Boundary markers: The skill documentation lacks instructions for implementing boundary markers or system-level warnings to ignore instructions embedded within the retrieved data.
  • Capability inventory: The skill provides the agent with write-access capabilities, including creating projects (GITLAB_CREATE_PROJECT), updating issues (GITLAB_UPDATE_PROJECT_ISSUE), and creating branches (GITLAB_CREATE_REPOSITORY_BRANCH).
  • Sanitization: There are no documented procedures for sanitizing or validating retrieved GitLab content before processing.
  • [EXTERNAL_DOWNLOADS]: The skill references and directs the user to connect a remote MCP server at https://rube.app/mcp. This is a standard integration for the Rube/Composio toolkit, which is a well-known service for extending AI agent capabilities through pre-built toolsets.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 18, 2026, 04:24 PM
Security Audit — agent-trust-hub — gitlab-automation