confluence-automation

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

• Third-party content exposure detected (medium risk: 0.30). In the “Search Content” workflow, the agent uses CONFLUENCE_SEARCH_CONTENT/CONFLUENCE_CQL_SEARCH to fetch Confluence pages and (optionally) hydrates full content via CONFLUENCE_GET_PAGE_BY_ID, so outsider-authored Confluence text can be ingested at runtime when a user prompts a search.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill requires connecting to the Rube MCP server and explicitly instructs fetching current tool schemas from https://rube.app/mcp at runtime (via RUBE_SEARCH_TOOLS), so content served from that URL can directly control agent tool schemas and behavior.