Compliance Control → Evaluate
AI-generatedSummary
Evaluate each control statement by classifying its security topic, scoring it using the control text and evidence URL count, and mapping detected topics to selected frameworks. The node also makes an authenticated POST request to its evaluation endpoint before processing input items; the submitted request body is fixed in the implementation.
Inputs
- Control Text — Policy or control statement to evaluate. Text keywords determine categories and contribute to the score.
- Evidence URLs — List of proof links, such as dashboards, reports, or configurations. The number and file or URL patterns affect scoring; no evidence adds a gap and a suggested action.
- Frameworks — Select frameworks for mapping detected categories to requirements. Defaults to Essential Eight, GDPR, ISO 27001, NIST CSF, PCI DSS, and SOC 2.
Output shape
One output item per input item, containing a JSON object.
Each object includes the input control text, detected categories, evidence URLs, status, score, confidence, evaluation, rationale, mapped requirements, selected frameworks, gaps, actions, and a notes string. The authenticated endpoint request uses a hard-coded framework, control, and empty evidence rather than the configured values. If that request fails, the operation errors before returning the per-item results.