Workflows

Multi-agent system

Fifteen specialist roles, typed artifact handoffs, activation rules, guardrails, model tiers, and architecture-impact adjudication.

Agents collaborate through validated artifacts—not hidden chain-of-thought. Every artifact records its agent definition version, model, context hash, evidence IDs, confidence, usage, latency, and decision summary.

Agents collaborate through typed artifactsSpecialists receive bounded context and hand off validated outputs instead of hidden reasoning.VERIOM FIELD GUIDEEVIDENCE FIRSTAgents collaborate through typed artifactsSpecialists receive bounded context and hand off validated outputs instead of hidden reasoning.TB1TB2COORDINATEANALYZEADJUDICATEactivatecontextartifactverified01Audit CoordinatorScope + stage02Context CuratorEvidence pack03Specialist agentsSecurity · cloud · CI/CD04Evidence VerifierClaims + contradictions05Report AdjudicatorCanonical resultNo hidden chain-of-thought is required to inspect a decision.
Agents collaborate through typed artifacts

Role registry

RolePrimary responsibilityTypical tier
Audit Coordinatorscope, activation, stage and retry decisionsLuna
Context Curatorassemble bounded ContextPacksLuna
Evidence Verifierreject unsupported claims and contradictionsLuna
Security Engineertechnical vulnerabilities and control failuresTerra
Security Architectboundaries, reachability, blast radius, structural weaknessTerra
Cloud/IaC Analystinfrastructure entities, policy, network, identityTerra
CI/CD Analystpipeline, artifact, deployment, and provenance relationshipsTerra
Threat Modelerassets, actors, paths, abuse cases, mitigationsTerra
Remediation Engineerdeterministic patch proposal and static rescan planTerra
Diagram Architectsafe views of the canonical graphTerra
Report Synthesizer / Adjudicatorreconcile artifacts into canonical JSONTerra
Finding Architecture Impact Adjudicatordetermine whether scanner observations materially affect evidenced architectureTerra
Compliance Control Analystmap evidence to bounded assurance questions without certifying complianceTerra
Architecture Diagnostic Analystidentify gaps, contradictions, staleness, and unlinked signalsLuna
Architecture History Analystexplain changes between comparable truth snapshotsLuna

Sol-class models are reserved for an explicitly approved deep review. They are never selected merely because another agent failed.

YAML definitions

Definitions under resources/llm/prompts, resources/llm/skills, and resources/llm/workflows in the Veriom API repository are versioned and validated at startup. Each definition declares:

  • role and activation rules;
  • model tier and bounded tools;
  • Pydantic input and output schemas;
  • token, latency, and cost limits;
  • memory policy and evidence requirements;
  • prompt-injection, secret, and unsupported-claim guardrails.

Unknown tools, invalid schemas, unsafe dynamic imports, or missing versions fail startup and CI.

Evidence adjudication

An observation becomes canonical only after the verifier can resolve every cited evidence ID in the same audit and workspace. Unsupported claims remain hypotheses. Conflicting agent artifacts are retained and sent to adjudication with the underlying evidence—not resolved by majority vote.

For scanner observations, citation validity is necessary but not sufficient. Architecture relevance requires a concrete affected entity, relationship, trust boundary, identity, pipeline, data flow, or constraint. An automatic exclusion requires scanner evidence plus independent structural evidence for a concrete exclusion. Missing deployment, cloud, runtime, or reachability evidence routes to human review.