github-automation

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). This skill explicitly calls GitHub APIs (e.g., GITHUB_SEARCH_CODE, GITHUB_GET_REPOSITORY_CONTENT, GITHUB_LIST_REPOSITORY_ISSUES, GITHUB_GET_A_PULL_REQUEST) in the SKILL.md "Search Code and Commits" and core workflow sections, which ingest public/user-generated GitHub content (issues, PRs, files) that the agent is expected to read and that can materially affect actions like merges or workflow dispatch.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). This skill requires connecting at runtime to the Rube MCP endpoint https://rube.app/mcp (via RUBE_SEARCH_TOOLS and RUBE_MANAGE_CONNECTIONS) to fetch tool schemas and auth flows that directly determine the agent's available prompts/instructions, making it a required external dependency.