datasources-provisioning

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

• Third-party content exposure detected (high risk: 0.85). Yes—at runtime the workflow fetches outsider-authored free text from public web endpoints (Grafana plugin catalog/CDN JSON like schema/dsconfig.json and v0alpha1.json) via curl, and that fetched prose/JSON is then used to construct the LLM’s context/output.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill performs runtime fetches of plugin schemas and example configs (e.g. https://plugins-cdn.grafana.net/<PLUGIN_ID>//public/plugins/<PLUGIN_ID>/schema/dsconfig.json and the grafana.com API endpoints), and those fetched JSON files directly control which fields/prompts the agent presents and how it builds the datasource config.