Manifold Grace

Research / Implementation programme

FUNCTIONAL DEMONSTRATION · DOCUMENTED SAMPLE DATA

Research Agent Lab

This interactive prototype performs local, deterministic comparison of two illustrative system architectures. It demonstrates the evidence schema, comparison logic and rule-based issue detection. It does not fetch live sources, operate autonomous AI agents, test suppliers or certify legal compliance. Everything shown is simulated educational test data, not findings about named products.

Comparison results

Choose two scenarios and run the comparison.
Sample system data and comparison rules

Each system has a stable ID, explicit component list, dependency edges, claimed controls, demonstration-only source record and authorisation level. Jaccard similarity is intersection over union for component sets, so 100% component overlap does not establish functional or ethical equivalence. Rules flag lack of approval, unknown logging and absence of versioned citations as review questions, not verified failures.

Sequential research-agent specification — 12 stages

StageRoleOutput / release gate
1Set purpose, jurisdiction and authorisationWritten scope and approved information sources
2Discover system candidatesDeduplicated inventory with source URLs
3Acquire authorised documentsVersioned originals, timestamps and hashes
4Extract architecture claimsComponents, edges, data flows and confidence
5Normalise conceptsControlled terminology and synonyms
6Verify each source and assertionClaim-to-evidence coverage and unresolved assertions
7Compare architecturesShared, distinct and structurally differing features
8Map tools and dependenciesLicences, providers, permissions, versions and known uncertainties
9Assess assurance requirementsApplicable duties, risks and evidence gaps; no automated certification
10Adversarially review findingsContradictions, false positives and counterexamples
11Human review and sign-offAccountable decision with corrections recorded
12Publish approved outputsVersioned report and audit trail; no automatic release

Engineering gates before real agents

Document parser fidelity; source permission and robots/site restrictions; scoped tool tokens; sandboxed read-only ingestion; privacy and retention; prompt-injection resilience; citation verification; withheld verdicts when evidence is inadequate; benchmark against human-labelled cases; approval before externally visible publication; monitoring, rollback and incident response.

Agent security: OWASP · Evidence provenance: W3C · Risk management: NIST · Systems Atlas