github-automation

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

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow uses Composio’s GitHub toolkit to call GitHub APIs (e.g., listing issues/PRs/workflows and fetching PR details), which can include outsider-authored free-text fields like issue/PR titles and bodies/comments that the agent ingests into its LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill requires connecting at runtime to the external MCP server https://rube.app/mcp to retrieve Rube tool schemas and tool behaviors (RUBE_SEARCH_TOOLS/RUBE_MANAGE_CONNECTIONS), which directly control the agent's available instructions and actions.