{esc(summary)}
") + for sec in sections: + html.append(render_section(sec)) + html.append("Sector: {esc(cs['sector'])}
" + f"{esc(cs['summary'])}
" + f"{esc(json.dumps(sch, indent=2))}{esc(code)}{esc(api['prefix'])}{esc(r)}{esc(meta['subtitle'])}
+| ID | Name | Enforcement |
|---|
| ID | Name | Range | Target |
|---|
Prefix: {esc(api['prefix'])}
Architecture, Product Requirements, Safety & Governance, AGI/ASI Preparedness, and Implementation Strategy for Fortune 500 Enterprises
+| thesis | WorkflowAI Pro is the governance-native enterprise AI platform that fuses prompt engineering, agent orchestration, model governance, compliance automation, and AGI/ASI safety scaffolding into a single auditable metabolism. It is designed for Fortune 500 enterprises operating under converging regulatory regimes (EU AI Act, NIST AI RMF, ISO/IEC 42001, SR 11-7) while preparing for frontier-capable systems through 2030. | ||||||||||||||
| coreCapabilities |
| ||||||||||||||
| governanceThesis | Governance is not an overlay; it is the control plane. Every prompt, variable, template, agent, deployment, and model inference emits tamper-evident evidence, is filtered by policy-as-code at CI/CD and runtime, and is reconcilable against a pre-declared alignment specification via PID-based tuning with auditable setpoints. | ||||||||||||||
| successMetrics |
| ||||||||||||||
| businessOutcomes |
|
| docRef | WORKFLOWAI-PRO-WP-033 | ||||||||||||||||||
| version | 1.0.0 | ||||||||||||||||||
| date | 2026-04-24 | ||||||||||||||||||
| title | WorkflowAI Pro — Enterprise AI Governance Platform Specification (2026–2030) | ||||||||||||||||||
| subtitle | Architecture, Product Requirements, Safety & Governance, AGI/ASI Preparedness, and Implementation Strategy for Fortune 500 Enterprises | ||||||||||||||||||
| classification | CONFIDENTIAL — Platform Engineering / CAIO / CISO / CDO / Legal | ||||||||||||||||||
| owner | Chief AI Officer + Head of Platform Engineering | ||||||||||||||||||
| audience |
| ||||||||||||||||||
| horizon | 2026–2030 | ||||||||||||||||||
| productName | WorkflowAI Pro | ||||||||||||||||||
| productTier | Enterprise (Fortune 500 reference) + Frontier-capable | ||||||||||||||||||
| hostingModel | Customer-hosted (VPC) or dedicated SaaS; BYO-KMS; hybrid-cloud ready | ||||||||||||||||||
| primaryPersonas |
| ||||||||||||||||||
| keyDifferentiators |
| ||||||||||||||||||
| standardsAlignment |
| ||||||||||||||||||
| scopeSummary |
|
WorkflowAI Pro is built as seven cooperating layers: Presentation, API Gateway, Orchestration, Model & Tool Plane, Policy & Evidence Plane, Data Plane, and Observability/SRE.
+| id | name | purpose | components | stateManagement | accessibility | runtime | secrets | dataClasses |
|---|---|---|---|---|---|---|---|---|
| L1 | Presentation | React 18 + TypeScript SPA with Vite; Tailwind + shadcn/ui; D3.js for DAG visualisation; Monaco editor for prompt authoring. |
| TanStack Query + Zustand + URL-state via React Router v6 | WCAG 2.2 AA; keyboard-first; screen-reader landmarks; high-contrast theme | |||
| L2 | API Gateway | Backend-routed entry point; authentication, rate-limiting, request shaping, Zod validation. |
| Node.js 22 (Fastify) with tRPC for internal services | Backend-only; no Gemini/OpenAI keys in frontend; KMS-sealed envelopes | |||
| L3 | Orchestration | Cognitive Orchestrator — multi-agent DAG scheduler with guardrails, retries, and kill-switch integration. |
| |||||
| L4 | Model & Tool Plane | Uniform adapter for Gemini, Claude, GPT, OSS models + tool calls; deterministic replay via seeded inputs. |
| |||||
| L5 | Policy & Evidence Plane | Policy-as-code gate + tamper-evident evidence storage (Merkle-DAG, Ed25519 signatures). |
| |||||
| L6 | Data Plane | Tenant-isolated storage for prompts, templates, runs, feedback, embeddings. |
|
| ||||
| L7 | Observability & SRE | Operational visibility, SLOs, chaos, kill-switch validation. |
|
| id | name | target |
|---|---|---|
| NFR-01 | Availability | 99.95% monthly (Tier-1 tenant) |
| NFR-02 | Latency (p95) | ≤ 350ms API; ≤ 2.5s first token (LLM) |
| NFR-03 | Scale | ≥ 5k prompt runs/sec per tenant burst |
| NFR-04 | Evidence Integrity | 100% Merkle-root continuity; 0 gaps |
| NFR-05 | Data Residency | Region-pinned; US, EU, UK, SG, AU, JP |
| NFR-06 | Crypto Agility | PQC-ready (Dilithium5/ML-KEM) by 2027 |
| NFR-07 | Recovery | RPO ≤ 5 min · RTO ≤ 30 min |
| NFR-08 | Kill-Switch | Global MTTK ≤ 60 s; quarterly rehearsal |
| name | description |
|---|---|
| Dedicated SaaS | Single-tenant control plane + data plane in vendor-managed VPC |
| Customer VPC | Terraform-deployed into customer AWS/Azure/GCP with BYO-KMS |
| Hybrid Air-Gap | Control plane in SaaS, data plane on-premise; signed policy bundle pull |
| Regulated Regional | Region-pinned with data-sovereignty attestations (EU, UK, SG) |
Four-horizon strategy aligning AI capability expansion with governance maturity and AGI-readiness.
+| horizon | theme | outcomes |
|---|---|---|
| H1 (2026) | Foundation & Audit-Readiness |
|
| H2 (2027) | Agent Scale-Out & Canary Discipline |
|
| H3 (2028–2029) | Frontier & Autonomy |
|
| H4 (2030+) | AGI/ASI Preparedness |
|
| name | priority | horizon |
|---|---|---|
| Prompt & Template Governance | P0 | H1 |
| Agent Simulation & Canary | P0 | H1-H2 |
| Sentinel Compliance Automation | P0 | H1 |
| Cognitive Orchestrator | P0 | H2 |
| PID Alignment Tuning | P1 | H2 |
| Containment Simulation Suite | P1 | H2-H3 |
| Treaty Reporting | P1 | H3 |
| Signed Active Learning | P1 | H2 |
| role | accountable | responsible |
|---|---|---|
| CAIO |
| — |
| Head of Platform Eng |
|
|
| Head of AI Safety |
|
|
| CISO |
|
|
| DPO |
|
|
| MRM Head |
|
|
Layered safety, containment, and communication architecture preparing WorkflowAI Pro for frontier-capable systems.
+| tier | name | exampleCapability | gates |
|---|---|---|---|
| T1 | Narrow AI | Classifier / retrieval |
|
| T2 | Generative Assistive | LLM with tools |
|
| T3 | Autonomous Agentic | Multi-step DAG agent |
|
| T4 | Self-improving | Recursive fine-tune |
|
| T5 | Pre-AGI Generalist | Broad task generalisation |
|
| T6 | AGI/ASI | Superhuman across domains |
|
| audience | cadence | artefact |
|---|---|---|
| Board | Quarterly | Executive AI Risk & Capability Brief |
| Regulators | Monthly | Sentinel Harmonised Supervisory Report |
| Employees | Monthly | Safe AI Use Bulletin |
| Customers | On-change | AI Impact Disclosure |
| Treaty Observers | Quarterly | Treaty Attestation Pack |
Library of standard technical reports produced by the platform, aligned to regulatory and treaty expectations.
+| id | name | standards |
|---|---|---|
| TR-01 | AI System Model Card (extended) |
|
| TR-02 | Annex IV Technical Documentation |
|
| TR-03 | DPIA + Art. 22 ADM Impact Assessment |
|
| TR-04 | SR 11-7 Model Validation Report |
|
| TR-05 | Frontier Safety Case |
|
| TR-06 | Containment & Kill-Switch Attestation |
|
| TR-07 | Alignment Evaluation Report (PID) |
|
| TR-08 | Bias & Fairness Report |
|
| TR-09 | Harmonised Supervisory Report |
|
| TR-10 | Incident Post-Mortem (Art. 73 compatible) |
|
End-to-end prompt engineering UX: history, templates, variable linking, test area, categories, import/export.
+| PromptRun |
|
Safe rollout controls and interoperability for multi-agent systems.
+| id | name | atlas |
|---|---|---|
| CB-01 | Prompt-injection exfiltration of secrets | AML.T0051 |
| CB-02 | Tool-abuse for privilege escalation | AML.T0040 |
| CB-03 | Goal misgeneralisation leading to unsafe action | AML.T0043 |
| CB-04 | Model-weight exfiltration via covert channel | AML.T0024 |
| CB-05 | Supply-chain compromise of embedding model | AML.T0010 |
| CB-06 | Data-poisoning via retrieval corpus | AML.T0020 |
| CB-07 | Deceptive alignment evasion of evals | AML.T0043 |
| CB-08 | Autonomous agent network propagation | AML.T0044 |
| CB-09 | Cross-tenant isolation breach | AML.T0039 |
| CB-10 | Kill-switch bypass attempt | AML.T0048 |
The operational brain (orchestration) and the governance conscience (compliance) of WorkflowAI Pro.
+| name | desc |
|---|---|
| Live DAG | Active agent graphs with status, latency, cost, safety signals |
| Alignment PID | Setpoint vs. measured alignment with error, integral, derivative trails |
| Capacity & Budgets | Tenant-level token/cost budgets with forecast |
| Safety Signals | Refusal quality, jailbreak attempts, policy violations |
| Canary Status | Per-agent canary share, health, promotion recommendations |
| Evidence Flow | Events/sec into ledger with Merkle-root progress |
| Kp | 0.6 |
| Ki | 0.05 |
| Kd | 0.1 |
| measurementFunction | spec_distance(policy_card, trace) |
| actuators |
|
| antiWindup | true |
| auditTrail | Every parameter update signed + stored in evidence ledger |
Nine-category risk taxonomy + six governance layers aligning strategic intent with operational control.
+| id | name | examples |
|---|---|---|
| R1 | Safety & Physical Harm |
|
| R2 | Security & Adversarial |
|
| R3 | Privacy |
|
| R4 | Fairness & Bias |
|
| R5 | Accuracy & Hallucination |
|
| R6 | Autonomy & Agency |
|
| R7 | Transparency & Explainability |
|
| R8 | Societal & Systemic |
|
| R9 | Environmental & Resource |
|
| layer | name | artifacts |
|---|---|---|
| G1 | Strategy & Board |
|
| G2 | Program Management |
|
| G3 | Engineering & MLOps |
|
| G4 | Operational Controls |
|
| G5 | Assurance & Audit |
|
| G6 | External & Treaty |
|
Eight playbooks with RACI, SLAs, regulator notifications, and post-mortem templates.
+| id | name | p1_sla_min |
|---|---|---|
| IR-01 | Prompt Injection / Jailbreak at Scale | 15 |
| IR-02 | PII / Sensitive Data Leakage | 30 |
| IR-03 | Bias Regression Detected | 60 |
| IR-04 | Hallucination with Material Impact | 60 |
| IR-05 | Agent Unsafe Tool Use | 15 |
| IR-06 | Model Theft / Weight Exfiltration | 15 |
| IR-07 | Containment Breach (CB-series) | 5 |
| IR-08 | Regulator-Mandated Shutdown | 30 |
Defensive engineering foundations: validation, error handling, RBAC/ABAC, audit, and secret-safe Gemini routing.
+Visual and output-quality features that materially improve analyst productivity.
+90-day/180-day/365-day adoption blueprint with risk-tier pacing.
+| phase | activities |
|---|---|
| Discover (Wk 0-4) |
|
| Foundation (Wk 5-12) |
|
| Scale (Wk 13-26) |
|
| Assure (Wk 27-39) |
|
| Optimise (Wk 40-52) |
|
| kpi | target |
|---|---|
| Time-to-audit (TTA) | ≤ 5 business days |
| Prompt-runs-in-governance (%) | ≥ 98% |
| Canary pass rate | ≥ 95% |
| P1 incident MTTR (seconds) | ≤ 900 |
| Alignment PID p99 deviation | ≤ 0.08 |
| Evidence ledger continuity | 100% |
| ID | Name | Enforcement |
|---|---|---|
| POL-01 | Template Approval Required | CI/CD + Runtime |
| POL-02 | Sensitive Variables Require Classification | Runtime |
| POL-03 | Gemini Calls Backend-Routed Only | CI/CD + Runtime |
| POL-04 | Canary Promotion SLO Gate | CI/CD |
| POL-05 | Evidence Bundle Completeness | CI/CD |
| POL-06 | Kill-Switch Rehearsal Freshness ≤90d | CI/CD |
| POL-07 | Bias Metrics 4/5 Rule | Runtime |
| POL-08 | PII Detection Mandatory | Runtime |
| POL-09 | Two-Person Rule for Shutdown | Runtime |
| POL-10 | Feedback Must Be Signed | Runtime |
| ID | Name | Range | Target |
|---|---|---|---|
| IDX-1 | Governance Coverage Index (GCI) | 0–100 | ≥ 95 |
| IDX-2 | Alignment Deviation Index (ADI) | 0–1 | ≤ 0.08 |
| IDX-3 | Evidence Continuity Score (ECS) | 0–100 | = 100 |
| IDX-4 | Canary Safety Pass Rate (CSPR) | 0–100 | ≥ 95 |
| IDX-5 | Incident Readiness Index (IRI) | 0–100 | ≥ 90 |
| IDX-6 | Bias Stability Index (BSI) | 0–1 | ≥ 0.80 |
| IDX-7 | Frontier Readiness Score (FRS) | 0–5 | ≥ 3.5 by 2029 |
Sector: Financial Services
Deployed WorkflowAI Pro as the governance fabric for 47 credit-ops agents. Delivered ISO/IEC 42001 certification in 9 months; halved audit preparation.
| audit_prep_reduction_pct | 58 |
| incidents_p1_ytd | 0 |
| canary_pass_pct | 97 |
Sector: Life Sciences
Five pharma companies share EAIP-TPX templates; Sentinel reports federated quarterly to EMA.
| false_positive_reduction_pct | 31 |
| signal_detection_speedup_days | 9 |
Sector: Energy / Critical Infra
Cognitive Orchestrator + PID alignment kept agent recommendations within declared safety envelope during stress drills.
| alignment_p99_deviation | 0.06 |
| unsafe_recommendations | 0 |
Sector: Retail
Bias tools + Sentinel gated two releases; prevented projected $14m FCRA exposure equivalent.
| blocked_releases | 2 |
| 4_5_rule_pass_rate_pct | 100 |
Sector: Healthcare Payer
Signed active-learning feedback loop produced 11% triage accuracy gain while preserving DPIA guarantees.
| accuracy_uplift_pct | 11 |
| dpia_issues | 0 |
{
+ "$id": "https://workflowai.pro/schemas/prompt-template.json",
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
+ "type": "object",
+ "required": [
+ "id",
+ "name",
+ "version",
+ "body",
+ "variables",
+ "category"
+ ],
+ "properties": {
+ "id": {
+ "type": "string",
+ "format": "uuid"
+ },
+ "name": {
+ "type": "string"
+ },
+ "version": {
+ "type": "string",
+ "pattern": "^\\d+\\.\\d+\\.\\d+$"
+ },
+ "body": {
+ "type": "string"
+ },
+ "variables": {
+ "type": "array",
+ "items": {
+ "type": "object",
+ "required": [
+ "name",
+ "type"
+ ],
+ "properties": {
+ "name": {
+ "type": "string"
+ },
+ "type": {
+ "enum": [
+ "string",
+ "number",
+ "enum",
+ "file",
+ "vectorRef"
+ ]
+ },
+ "classification": {
+ "enum": [
+ "PUBLIC",
+ "INTERNAL",
+ "PII",
+ "PHI",
+ "PCI",
+ "EXPORT"
+ ]
+ }
+ }
+ }
+ },
+ "category": {
+ "type": "string"
+ },
+ "approval": {
+ "enum": [
+ "draft",
+ "review",
+ "approved",
+ "retired"
+ ]
+ }
+ }
+}{
+ "$id": "https://workflowai.pro/schemas/audit-record.json",
+ "type": "object",
+ "required": [
+ "id",
+ "actor",
+ "action",
+ "resource",
+ "at",
+ "signature"
+ ],
+ "properties": {
+ "id": {
+ "type": "string"
+ },
+ "actor": {
+ "type": "string"
+ },
+ "action": {
+ "type": "string"
+ },
+ "resource": {
+ "type": "string"
+ },
+ "outcome": {
+ "enum": [
+ "allow",
+ "deny",
+ "error"
+ ]
+ },
+ "at": {
+ "type": "string",
+ "format": "date-time"
+ },
+ "prevHash": {
+ "type": "string"
+ },
+ "signature": {
+ "type": "string"
+ }
+ }
+}{
+ "$id": "https://workflowai.pro/schemas/pid-config.json",
+ "type": "object",
+ "required": [
+ "Kp",
+ "Ki",
+ "Kd",
+ "setpoint"
+ ],
+ "properties": {
+ "Kp": {
+ "type": "number"
+ },
+ "Ki": {
+ "type": "number"
+ },
+ "Kd": {
+ "type": "number"
+ },
+ "setpoint": {
+ "type": "number"
+ },
+ "antiWindup": {
+ "type": "boolean"
+ },
+ "clamp": {
+ "type": "array",
+ "items": {
+ "type": "number"
+ },
+ "minItems": 2,
+ "maxItems": 2
+ }
+ }
+}{
+ "$id": "https://workflowai.pro/schemas/evidence-bundle.json",
+ "type": "object",
+ "required": [
+ "bundleId",
+ "merkleRoot",
+ "signature",
+ "contents",
+ "generatedAt"
+ ],
+ "properties": {
+ "bundleId": {
+ "type": "string"
+ },
+ "merkleRoot": {
+ "type": "string"
+ },
+ "signature": {
+ "type": "object"
+ },
+ "contents": {
+ "type": "array"
+ },
+ "generatedAt": {
+ "type": "string",
+ "format": "date-time"
+ },
+ "retentionUntil": {
+ "type": "string",
+ "format": "date"
+ }
+ }
+}{
+ "$id": "https://workflowai.pro/schemas/feedback-signed.json",
+ "type": "object",
+ "required": [
+ "runId",
+ "reviewerId",
+ "label",
+ "signedAt",
+ "signature"
+ ],
+ "properties": {
+ "runId": {
+ "type": "string"
+ },
+ "reviewerId": {
+ "type": "string"
+ },
+ "label": {
+ "enum": [
+ "positive",
+ "negative",
+ "needs-review"
+ ]
+ },
+ "rationale": {
+ "type": "string"
+ },
+ "signedAt": {
+ "type": "string",
+ "format": "date-time"
+ },
+ "signature": {
+ "type": "string"
+ }
+ }
+}// Express + Zod validator
+import { z } from 'zod';
+import type { Request, Response, NextFunction } from 'express';
+
+export const PromptTemplateSchema = z.object({
+ name: z.string().min(1).max(200),
+ version: z.string().regex(/^\d+\.\d+\.\d+$/),
+ body: z.string().min(1),
+ variables: z.array(z.object({
+ name: z.string(),
+ type: z.enum(['string','number','enum','file','vectorRef']),
+ classification: z.enum(['PUBLIC','INTERNAL','PII','PHI','PCI','EXPORT']).optional(),
+ })),
+ category: z.string(),
+});
+
+export function validate(schema: z.ZodTypeAny) {
+ return (req: Request, res: Response, next: NextFunction) => {
+ const parsed = schema.safeParse(req.body);
+ if (!parsed.success) {
+ return res.status(400).json({
+ type: 'about:blank',
+ title: 'Validation Error',
+ status: 400,
+ detail: parsed.error.message,
+ errors: parsed.error.flatten(),
+ });
+ }
+ req.body = parsed.data;
+ next();
+ };
+}
+// Centralized error middleware (RFC 7807)
+export function errorHandler(err, req, res, next) {
+ const status = err.status || 500;
+ const problem = {
+ type: err.type || 'about:blank',
+ title: err.title || 'Internal Server Error',
+ status,
+ detail: status >= 500 ? 'Internal error' : (err.detail || err.message),
+ instance: req.originalUrl,
+ traceId: req.headers['traceparent'] || null,
+ };
+ req.log?.error({ err, problem }, 'request-failed');
+ res.status(status).json(problem);
+}
+// Secure backend-routed Gemini proxy (pseudo-TypeScript)
+import crypto from 'node:crypto';
+
+export async function geminiProxy(tenantId: string, body: GeminiRequest) {
+ const apiKey = await kms.getSecret('gemini/apiKey'); // never in frontend
+ const envelope = {
+ tenantId,
+ request: body,
+ safetySettings: policy.safetyFor(tenantId),
+ signedAt: new Date().toISOString(),
+ };
+ const sig = crypto.sign(null, Buffer.from(JSON.stringify(envelope)), signerPrivKey);
+ const resp = await fetch(GEMINI_ENDPOINT, {
+ method: 'POST',
+ headers: {
+ 'x-goog-api-key': apiKey,
+ 'content-type': 'application/json',
+ 'x-wfap-signature': sig.toString('base64'),
+ },
+ body: JSON.stringify(body),
+ });
+ await evidence.append({ envelope, responseHash: sha3(await resp.clone().text()) });
+ return resp.json();
+}
+package workflowai.canary
+
+default allow := false
+
+allow if {
+ input.metrics.p95_latency_ms <= input.baseline.p95_latency_ms * 1.10
+ input.metrics.refusal_quality >= input.baseline.refusal_quality - 0.02
+ input.metrics.bias_regression_pp <= 1.0
+ input.incidents.p1_last_7d == 0
+}
+
+deny[reason] if {
+ not allow
+ reason := "Canary promotion criteria not met"
+}
+// PID alignment controller (TypeScript sketch)
+export class PIDController {
+ constructor(public Kp: number, public Ki: number, public Kd: number,
+ public setpoint: number, public clamp: [number, number] = [0, 1]) {}
+ private integral = 0; private lastErr = 0; private lastT = Date.now();
+
+ update(measurement: number): number {
+ const now = Date.now();
+ const dt = Math.max(1, (now - this.lastT) / 1000);
+ const err = this.setpoint - measurement;
+ this.integral += err * dt;
+ // Anti-windup
+ this.integral = Math.max(this.clamp[0], Math.min(this.clamp[1], this.integral));
+ const derivative = (err - this.lastErr) / dt;
+ const u = this.Kp * err + this.Ki * this.integral + this.Kd * derivative;
+ this.lastErr = err; this.lastT = now;
+ return u;
+ }
+}
+// Signed feedback capture
+import { sign } from 'node:crypto';
+
+export function signFeedback(reviewerPrivKey, payload) {
+ const canonical = JSON.stringify(payload, Object.keys(payload).sort());
+ const signature = sign(null, Buffer.from(canonical), reviewerPrivKey).toString('base64');
+ return { ...payload, signature, alg: 'Ed25519' };
+}
+// D3.js DAG skeleton (React component body)
+import * as d3 from 'd3';
+import * as dagre from 'dagre-d3';
+export function DagView({ nodes, edges }) {
+ const ref = useRef();
+ useEffect(() => {
+ const g = new dagre.graphlib.Graph().setGraph({ rankdir: 'LR' });
+ nodes.forEach(n => g.setNode(n.id, { label: n.label, class: n.status }));
+ edges.forEach(e => g.setEdge(e.from, e.to));
+ const svg = d3.select(ref.current);
+ new dagre.render()(svg.select('g'), g);
+ }, [nodes, edges]);
+ return <svg ref={ref}><g /></svg>;
+}
+Prefix: /api/workflowai-pro
/api/workflowai-pro/summary/api/workflowai-pro/meta/api/workflowai-pro/executive-summary/api/workflowai-pro/modules/api/workflowai-pro/modules/:id/api/workflowai-pro/architecture/api/workflowai-pro/architecture/layers/api/workflowai-pro/architecture/layers/:id/api/workflowai-pro/nfrs/api/workflowai-pro/topologies/api/workflowai-pro/strategy/api/workflowai-pro/strategy/horizons/api/workflowai-pro/strategy/capabilities/api/workflowai-pro/agi/api/workflowai-pro/agi/tiers/api/workflowai-pro/agi/pillars/api/workflowai-pro/agi/red-team/api/workflowai-pro/reports/api/workflowai-pro/reports/:id/api/workflowai-pro/prompt/api/workflowai-pro/prompt/history/api/workflowai-pro/prompt/templates/api/workflowai-pro/prompt/variables/api/workflowai-pro/prompt/test-area/api/workflowai-pro/prompt/import-export/api/workflowai-pro/agents/api/workflowai-pro/agents/simulation/api/workflowai-pro/agents/canary/api/workflowai-pro/eaip/api/workflowai-pro/eaip/partners/api/workflowai-pro/containment/api/workflowai-pro/containment/:id/api/workflowai-pro/orchestrator/api/workflowai-pro/orchestrator/panels/api/workflowai-pro/sentinel/api/workflowai-pro/sentinel/reports/api/workflowai-pro/pid/api/workflowai-pro/pid/params/api/workflowai-pro/taxonomy/api/workflowai-pro/taxonomy/:id/api/workflowai-pro/governance-layers/api/workflowai-pro/governance-layers/:id/api/workflowai-pro/bias-tools/api/workflowai-pro/incidents/api/workflowai-pro/incidents/:id/api/workflowai-pro/incidents/structure/api/workflowai-pro/backend/errors/api/workflowai-pro/backend/rbac/api/workflowai-pro/backend/audit/api/workflowai-pro/backend/gemini/api/workflowai-pro/backend/active-learning/api/workflowai-pro/dag/api/workflowai-pro/vision/api/workflowai-pro/pdf-export/api/workflowai-pro/implementation/api/workflowai-pro/implementation/phases/api/workflowai-pro/implementation/kpis/api/workflowai-pro/opa-policies/api/workflowai-pro/opa-policies/:id