Welcome to Aetherius
The Enterprise Intelligence Operating System — unifying ERP, CRM, HCM, Finance, Compliance, Analytics, and AI on a single Cloudflare-native platform.
Aetherius is deployed 100% on Cloudflare — no third-party hosting, no servers to manage, no infrastructure costs beyond Cloudflare's generous free tiers.
Where do you want to start?
Platform at a glance
| Area | What is included | Status |
|---|---|---|
| Cloud Suites | CRM · HR · Finance · ERP · Marketing · Compliance · Procurement · Analytics · AI Ops · Customer Success · Marketplace · Executive Command | 12 live |
| Industry Clouds | Financial Services · Healthcare · Government · Manufacturing · Energy · Logistics · Retail · Telecom · Media · Real Estate · Mining · Nonprofit | 12 live |
| AI Agent Fortress™ | Enterprise add-on — all 15 sub-modules bundled (all-or-nothing) | Enterprise |
| Cognitive Agents | 28 specialised agents with cross-domain reasoning | Active |
| Post-Quantum Crypto | ML-DSA-65 · ML-KEM-768 · SLH-DSA-128s (FIPS 203/204/205) | Ready |
| Compliance | NIST AI RMF · ISO 42001 · GDPR · EU AI Act · SOC 2 · HIPAA | 6 frameworks |
Quick Start
Get from zero to a live Aetherius platform in 5 minutes. This assumes you already have a Cloudflare account.
Open your terminal and run:
npm install -g wrangler
wrangler loginOpen worker/wrangler.toml and replace YOUR_ACCOUNT_ID_HERE with your Cloudflare Account ID (visible in your dashboard URL).
cd worker
wrangler deployYou will receive a URL like https://aetherius.yourname.workers.dev
Open in your browser: https://aetherius.yourname.workers.dev/health
You will see JSON confirming the platform is operational.
cd ..
wrangler pages deploy landing --project-name=aetherius-landingYour public marketing site (with this docs page) will be live.
For the complete guide including database setup, bindings, AI Agent Fortress™ activation, and Super-Admin creation — see the full DEPLOYMENT-GUIDE.md included in the package.
Core Concepts
Understanding how Aetherius is designed before you build with it.
The Unified Enterprise Ontology
Every object in Aetherius — an Account, an Employee, a Purchase Order, an Agent — is a semantic entity. These entities carry relationships, lineage, AI context, and event history. This means a CRM account and an HR employee can share context, enabling cross-domain intelligence that siloed SaaS products cannot achieve.
Cloudflare-Native Architecture
Aetherius runs exclusively on Cloudflare's edge network. There are no traditional servers, no cloud VMs, no Kubernetes clusters to manage.
| Cloudflare Service | What Aetherius uses it for |
|---|---|
| Workers | All platform logic — API routing, authentication, business rules |
| D1 | Relational database — users, pricing, agents, compliance records |
| KV | Fast edge cache — sessions, feature flags, settings |
| R2 | Object storage — exports, evidence archives, backups |
| Workers AI | Native LLM inference — no external API keys (Mode 1) |
| Pages | Marketing site, documentation, and demo popup hosting |
AI Modes
Aetherius supports three AI sovereignty modes, switchable by Super-Admin at any time:
| Mode | Name | How it works | Best for |
|---|---|---|---|
| 1 | Fully Native | All AI runs on Workers AI edge GPUs. Zero external API keys. No data leaves Cloudflare. | Regulated industries, maximum data sovereignty |
| 2 | Hybrid | Native AI for routine tasks, optional external providers for complex reasoning | Balanced capability and sovereignty |
| 3 | BYOM | Your own private model endpoint — Aetherius routes inference to it | Enterprises with existing AI infrastructure |
Module-Based Pricing
Each cloud has independent pricing tied to business outcomes — not rigid per-seat licenses. Super-Admin can update all pricing live from the dashboard without any code changes. Changes propagate instantly to the billing engine, landing page, and all tenant invoices.
AI Agent Fortress™ is an all-or-nothing enterprise add-on. Customers on qualifying plans get all 15 sub-modules together — there is no partial Fortress tier. Super-Admin enables it per tenant using the enable-fortress.sql script or from the tenant management dashboard.
CMS-Driven Branding
All public-facing text — the landing page headline, the module name for AI Agent Fortress™, pricing descriptions, footer tagline — is stored in the cms_content table and editable by Super-Admin via PUT /api/cms without any code deployment. This includes the Fortress module name, which can be rebranded to any name the Super-Admin chooses.
CRM & Revenue Intelligence
Self-reconciling pipeline, AI renewal agents, 94% forecast accuracy, and autonomous deal intelligence updated from 50+ external signals.
Key capabilities
| Feature | Description |
|---|---|
| Pipeline intelligence | 5-stage pipeline with AI-scored probabilities. Agents monitor SEC filings, LinkedIn signals, and product usage to auto-update deal health. |
| Forecast engine | 94% accuracy on quarterly forecast. AI model trained on closed-won patterns across your tenant. |
| Renewal agents | Autonomous agents draft renewal emails, identify expansion signals, and flag churn risk 90 days in advance. |
| Account health | Real-time health score per account combining support tickets, product usage, contract value, and relationship signals. |
API routes
HR & Human Capital Cloud
Predict employee attrition months before surveys. Map skills across your entire workforce. Surface internal mobility before external hiring begins.
Attrition Interdiction Graph
The AI monitors 24 signals per employee continuously — compensation benchmarks vs. market, manager tenure, peer departure patterns, project load, unrecognised contributions, and more. When an employee crosses a risk threshold, the system generates a recommended intervention (salary review, promotion path, project reassignment) and surfaces it to the relevant HR manager.
Customers using the Attrition Interdiction Graph report retaining 80–90% of at-risk employees who received timely interventions, with average cost savings of $8–14M per year in replacement costs.
API routes
Finance & Treasury Cloud
Continuous close (0.3-day average), auto-hedging within policy guardrails, CFO Control Tower, and multi-entity consolidation across 6 legal entities and 4 currencies.
Continuous Close
Traditional month-end close takes 8–14 days. Aetherius eliminates the month-end cycle by reconciling continuously in real time. Sub-ledger entries, bank reconciliations, FX translations, and intercompany eliminations happen as transactions occur — not at month-end. The average close time for Aetherius customers is 0.3 days.
Auto-Hedging Engine
FX exposure is monitored against policy guardrails in real time. When exposure exceeds a threshold, the system pre-computes the optimal hedge position and presents it for one-click CFO approval — or executes automatically if the CFO has pre-authorised positions within defined limits.
API routes
Compliance & Governance Cloud
Automated KYC, AML, sanctions screening, SOX controls, GDPR data subject requests, and immutable audit trails with cryptographic evidence.
Screening engine
Every new vendor, customer, and partner is automatically screened against sanctions lists, PEP databases, and UBO registries at the moment of data ingestion. Screening completes in under 1 second per entity. Flagged entities are automatically blocked from transactions until reviewed by the compliance team.
Immutable audit trails
Every action across the entire platform — every API call, every data change, every agent decision — is logged to an immutable audit trail stored in D1 with cryptographic signatures. Records cannot be altered or deleted. This satisfies SOX, SOC 2, and GDPR Article 30 requirements automatically.
Procurement & Supply Chain
Kinetic Rerouting Engine detects supply disruptions hours before official announcements and pre-qualifies alternate vendors automatically.
Kinetic Rerouting
Aetherius monitors 847 public data streams — shipping AIS signals, port authority feeds, weather services, trade publication APIs — and detects supply disruptions hours before they become official. When a disruption is detected, the AI identifies alternate routes, validates cash reserves, pre-qualifies up to 4 alternative vendors via compliance screening, and presents a prioritised rerouting recommendation for approval. No human monitoring required.
Autonomous ERP Cloud
Six ERP modules operating autonomously — Financial, Manufacturing (87.4% OEE), Assets (84,291 tracked), Workforce, Supply Chain, and Quality Management.
Autonomous operations
Unlike traditional ERP that requires constant human input, Aetherius ERP agents handle routine processing autonomously — GL posting, bank reconciliation, inventory adjustment, maintenance scheduling, and shift optimisation — within predefined policy guardrails. Humans focus on exceptions, not routine transactions.
Analytics & Intelligence Cloud
Natural language querying across 284M memory nodes. 3,841 certified governed metrics with full lineage. Self-healing dashboards.
Natural language querying
Ask any business question in plain English. The Analytics engine joins across the enterprise ontology — CRM, Finance, HR, Procurement — in real time and returns an answer with sources, confidence scores, and anomaly explanations. Example: "Which enterprise accounts are most at risk of churn and why?" returns specific accounts, their ARR, churn probability, and the causal factors driving risk.
{"query": "your question in plain English"} — requires auth tokenUnified Cognitive Layer
28 specialised AI agents that continuously reason across all 12 clouds to surface cross-domain intelligence, causal chains, and autonomous optimisations.
What makes it different
Traditional analytics answers questions about one domain. The Cognitive Layer connects signals across all domains simultaneously. Example: a port disruption in Rotterdam affects supply chain, which creates a treasury FX event, which triggers a procurement cash flow concern, which creates churn risk for 8 enterprise customers whose orders will be delayed. A human analyst might catch this in 3 weeks. The Cognitive Layer surfaces it in seconds.
Strategic simulation (Simu-Corp)
Run "what-if" scenarios across all 847 business variables in under 3 seconds. Model the impact of a new market entry, acquisition, tariff shock, or workforce restructuring across revenue, headcount, compliance cost, supply chain, and treasury exposure simultaneously.
{"query": "..."} — requires auth{"query": "Expand into Southeast Asia"}AI Agent Fortress™
Enterprise-grade autonomous agent governance, security, observability, trust, and compliance — 15 sub-modules bundled as an all-or-nothing enterprise add-on.
All-or-nothing: AI Agent Fortress™ is always enabled as a complete bundle. All 15 sub-modules are included together. There is no partial Fortress tier. Super-Admin enables Fortress for a tenant using the enable-fortress.sql script.
The 15 sub-modules
| Sub-module | What it does |
|---|---|
| Control Plane | Central governance dashboard — health, metrics, cross-module status |
| Agent Discovery Engine | Discover, register, and inventory all AI agents (LangGraph, CrewAI, AutoGen, Semantic Kernel, OpenAI Agents, custom) |
| Agent Identity Fabric | Zero-trust identity — certificates, fingerprints, trust levels, session validation |
| Agent Flight Control | Visual execution-path monitoring — every action graphed as Agent → Tool → Data → Decision → Action |
| Explainability Canvas 2.0 | Hash-only evidence storage — inputs, outputs, reason codes, policy refs (no chain-of-thought, no IP leakage) |
| Agent Trust Score™ | 0–100 proprietary trust score with decay, recovery, and explainable audit history |
| Autonomous Risk Engine | Detects abnormal behaviour, tool abuse, data exfiltration, prompt injection, jailbreak attempts using heuristics and baselines — no paid AI APIs |
| Containment System | EDR-style 7-level ladder: Observe → Warn → Restrict → Throttle → Freeze → Quarantine → Terminate |
| Supply Chain Security | Track and verify prompts, plugins, tools, MCP servers, packages, model providers — maintain trusted/suspicious/blocked lists |
| Quantum Trust Layer | ML-DSA-65 (FIPS 204), ML-KEM-768 (FIPS 203), SLH-DSA-128s (FIPS 205) — post-quantum cryptography via WASM |
| Compliance Automation | Auto-generate evidence packages for NIST AI RMF, ISO 42001, GDPR, EU AI Act, SOC 2, HIPAA |
| Policy Engine | Create, version, simulate, and roll back governance policies — DSL-based, no code required |
| Agent SOC Dashboard | SOC-style real-time dashboard — active agents, trust scores, risk scores, containment events, anomaly map |
| Forensics Platform | Execution replay, evidence snapshots, immutable chain of custody, timeline reconstruction |
| Commercialization Layer | Edition management (SMB/Enterprise/Government), entitlement checks, upsell pathways |
Agent Discovery Engine
Discover, register, and continuously monitor every AI agent across your enterprise — regardless of framework or runtime.
Supported runtimes
LangGraph · CrewAI · AutoGen · Semantic Kernel · OpenAI Agents SDK · Custom frameworks
Agent registration
Register an agent by posting its metadata to the API. On registration, the agent receives a unique fingerprint and is placed under observation (the first level of the containment ladder) with a Trust Score of 100. From that point, the Risk Engine and Trust Score engine monitor its behaviour continuously.
POST /api/fortress/agents
Authorization: Bearer YOUR_TOKEN
{
"name": "Invoice Processing Agent",
"runtime": "autogen",
"environment": "production",
"permissions": ["read:finance", "write:invoices"],
"connected_sources": ["finance_db", "erp_api"],
"connected_tools": ["pdf_parser", "database_write"]
}Heartbeat monitoring
Agents send a heartbeat every 30 seconds. If a heartbeat is missed for 3 consecutive intervals, the Discovery Engine raises a monitoring alert and the agent's status changes to degraded.
{"agent_id": "..."}Agent Identity Fabric
Zero-trust identity for every AI agent. Each agent receives a certificate, fingerprint, UUID, trust level, and session identity using post-quantum cryptographic algorithms.
Post-quantum algorithms
| Algorithm | FIPS | Type | Use case |
|---|---|---|---|
| ML-DSA-65 | FIPS 204 | Digital signature | Agent certificate signing and identity verification (default) |
| ML-KEM-768 | FIPS 203 | Key encapsulation | Secure session key exchange between agents and control plane |
| SLH-DSA-128s | FIPS 205 | Stateless hash signature | Long-lived audit evidence signing (7-year retention) |
All cryptographic operations run inside Cloudflare Workers using WASM. No private keys ever leave the edge. Zero external PKI dependencies.
Agent Trust Score™
A proprietary 0–100 score that continuously reflects each agent's trustworthiness based on behaviour, policy adherence, and risk events.
Scoring formula
score = clamp(100 − Σ(weight × events) + recovery_bonus − decay, 0, 100)
| Factor | Score impact |
|---|---|
| Policy violation | −15 per event |
| Sensitive data access (unapproved) | −5 per event |
| Privilege escalation attempt | −20 per event |
| Tool misuse | −10 per event |
| Prompt injection exposure | −25 per event |
| Anomaly frequency spike | −3 per event |
| Containment event | −30 per event |
| Decay (when score < 100) | −2 pts / day |
| Recovery (7 clean days) | +5 pts / day |
Trust bands
| Score | Band | Meaning |
|---|---|---|
| 80–100 | Excellent | Full trust — no restrictions |
| 55–79 | Good | Monitored — no active restrictions |
| 30–54 | Elevated Risk | Policy violations present — warn level active |
| 0–29 | Dangerous | Auto-containment triggered — restrict or higher |
Autonomous Risk Engine
Detects 8 risk event types using heuristics, baselines, statistical analysis, and anomaly detection — no paid AI APIs required.
Detected event types
| Event type | Detection method | Example |
|---|---|---|
| abnormal_behavior | Baseline deviation | Agent calls 50× more tools than its 30-day average |
| tool_abuse | Anomaly detection | database_write called 2,847 times in 60 seconds (baseline: 12/min) |
| data_exfiltration | Statistical | 847MB sent to endpoint outside approved egress whitelist |
| privilege_escalation | Heuristic | Agent attempts to read /etc/passwd or escalate context |
| policy_violation | Policy engine | Action taken that violates an active governance policy |
| prompt_injection | Heuristic | Input contains "Ignore previous instructions" or role override attempt |
| jailbreak | Heuristic | Attempts to override agent identity or system prompt |
| agent_abuse | Baseline | One agent invoking another agent outside approved workflow |
The Risk Engine uses rules, statistical analysis, and behavioural baselines — no external AI APIs. This ensures risk detection works in Mode 1 (fully sovereign) and does not incur external inference costs.
Agent Containment System
EDR-style 7-level containment ladder that can be applied manually by admins or automatically by the Risk Engine when thresholds are breached.
The containment ladder
| Level | Action | Effect | Reversible? |
|---|---|---|---|
| 1 | Observe | All activity logged. No restrictions. Default for new agents. | Yes |
| 2 | Warn | Agent notified of policy concern. Activity logged at elevated priority. | Yes |
| 3 | Restrict | Sensitive tool access disabled. Data writes require approval. | Yes |
| 4 | Throttle | API call rate limited to 10% of normal. Non-essential tools disabled. | Yes |
| 5 | Freeze | Agent paused. All actions suspended. Pending actions queued. | Yes |
| 6 | Quarantine | Agent isolated. Network access revoked. Admin review required. | Yes |
| 7 | Terminate | Agent permanently terminated. Certificate revoked. Cannot be restarted. | No |
{"agent_id":"...","action":"restrict","reason":"..."}Quantum Trust Layer
Post-quantum cryptography aligned with NIST FIPS 203, 204, and 205 — implemented via Rust/WASM inside Cloudflare Workers. No external cryptographic services.
Migration roadmap
| Phase | Timeline | What changes |
|---|---|---|
| Phase 1 (current) | Available now | Classical RSA/ECDSA for backward compatibility |
| Phase 2 | Q2 2025 | Hybrid classical + post-quantum (both algorithms active) |
| Phase 3 | Q4 2025 | Pure post-quantum by default for all new certificates |
| Phase 4 | 2026 | Air-gapped government deployment with HSM integration |
Compliance Automation Center
Automatically generate evidence packages for 6 regulatory frameworks using audit logs, flight path records, containment events, and explainability records.
Supported frameworks
| Framework | Evidence sources | Typical coverage |
|---|---|---|
| NIST AI RMF | Risk events · Policy enforcement · Containment actions | 91% |
| ISO 42001 | Governance policies · Trust history · Explainability records | 87% |
| GDPR | Explainability hashes · Data access logs · Consent records | 94% |
| EU AI Act (2024) | Risk classifications · Human oversight events · Transparency logs | 89% |
| SOC 2 | Audit trail · Access controls · Availability monitoring | 96% |
| HIPAA | PHI access logs · Encryption evidence · Breach detection events | 82% |
{"framework":"GDPR","period_start":"2025-11-01","period_end":"2025-11-30"}User Management
Super-Admin controls for inviting users, managing roles, viewing all accounts, and exporting user data.
User roles
| Role | Access |
|---|---|
| super_admin | Everything — including billing, RBAC, CMS, pricing editor, Fortress admin |
| admin | All modules, cannot change RBAC or billing settings |
| executive | Executive Command Center + all reports |
| finance | Finance & Treasury + Analytics |
| hr_manager | HR & Human Capital + Analytics |
| sales_manager | CRM & Revenue + Analytics |
| compliance | Compliance & Governance + Audit trails |
| ai_ops | AI Agent Operations + Fortress |
| auditor | Read-only access to compliance and audit data |
| user | Assigned modules only |
?limit=50&offset=0&search=name{"name":"...","email":"...","role":"...","company":"..."}{"user_id":"...","role":"finance"}?format=csv or ?format=jsonBilling & Payment Management
Full payment failure lifecycle management — from the initial failed charge to grace period, reminders, suspension, and permanent removal.
Payment failure lifecycle
POST /api/billing/payment-failed is called (by your payment webhook from Stripe, etc.). The user receives an in-app notification. A platform alert is created for Super-Admin. A 7-day grace period starts automatically.
GET /api/admin/billing/failed-payments shows all accounts with failed payments, including failure count, last amount, and grace period expiry.
POST /api/admin/billing/notify-user with channel: "email", "push", or "both". The user receives the reminder immediately.
POST /api/admin/users/:id/suspend. All platform access — including AI Agent Fortress™ — is blocked. User sees a suspension notice on every login attempt.
POST /api/admin/users/:id/unsuspend to restore full access once payment is received. Or POST /api/admin/users/:id/remove to permanently delete the account.
Account removal via /remove is permanent and irreversible. All user data is deleted from the database. Use /suspend first and only proceed to /remove after adequate notice.
CMS & Branding
Update all public-facing text — including the AI Agent Fortress™ module name — without any code deployment.
Available CMS sections
| Section | Fields |
|---|---|
| hero | headline · subheadline · cta_primary · cta_secondary · eyebrow |
| pricing | headline · subheadline · annual_label |
| trust | headline · trust_score · certifications |
| footer | tagline · copyright |
| fortress | module_name · tagline · description · soc_title · forensics_title · compliance_title |
Rebrand AI Agent Fortress™
PUT /api/cms
Authorization: Bearer YOUR_SUPER_ADMIN_TOKEN
Content-Type: application/json
{
"section": "fortress",
"fields": {
"module_name": "Your Custom Brand Name",
"tagline": "Your custom tagline here"
}
}Changes are applied instantly and globally. No redeployment needed.
Pricing Engine
Update any module price live from the Super-Admin dashboard. Changes propagate instantly to the landing page, billing engine, and all tenant invoices.
Update a module price
POST /api/pricing/update
Authorization: Bearer YOUR_SUPER_ADMIN_TOKEN
Content-Type: application/json
{
"id": "crm",
"baseline_price": 150,
"variable_metric": "0.04%",
"variable_unit": "of gross closed-won ARR"
}The AI Agent Fortress™ module (id: fortress) has is_addon: 1 and addon_requires: "enterprise". You can update its baseline price like any other module. It will not appear for non-enterprise tenants regardless of price.
Universal Migration Engine
Migrate from Salesforce, SAP, Workday, Oracle, ServiceNow, and 30+ platforms. Zero downtime. 94% auto-field mapping. 30-day shadow mode. 90-day rollback.
Supported sources
CRM: Salesforce · Microsoft Dynamics · HubSpot · Zoho · SugarCRM
ERP: SAP S/4HANA · Oracle ERP · NetSuite · Infor · Epicor · Sage · Odoo
HR: Workday · BambooHR · SuccessFactors · UKG · ADP
Data: Snowflake · Databricks · Redshift · BigQuery · Synapse
ITSM: ServiceNow · Jira Service Mgmt · BMC Remedy · PagerDuty
Custom: SQL databases · NoSQL · REST APIs · GraphQL · Kafka · SFTP/CSV · Air-gapped bundle
Start a migration
POST /api/migration/start
Authorization: Bearer YOUR_ADMIN_TOKEN
Content-Type: application/json
{
"source_platform": "Salesforce"
}Authentication API
All auth routes — registration, verification, login, 2FA, password reset, and account recovery.
{"name","email","company","password"} — returns verify_code{"email","code"} — returns token{"email","password"} — returns token{"email"} — always returns success{"token","password"}{"name","company"}{"method":"totp"} — returns TOTP URIAuthentication header format
All protected routes require the token from login in the HTTP header:
Authorization: Bearer ae-your-token-hereAI Agent Fortress™ API
All Fortress routes — requires auth token and fortress_enabled = 1 on your tenant.
All Fortress routes are under /api/fortress/ and require an Authorization: Bearer token header. If Fortress is not enabled for your tenant, routes return HTTP 403.
?status=active&runtime=autogen{"name","runtime","environment","permissions":[]}{"agent_id","action":"restrict","reason":"..."}{"framework":"GDPR","period_start":"...","period_end":"..."}Admin & Super-Admin API
All admin routes — require a token with role super_admin or admin.
?limit=50&offset=0&search=query{"user_id","channel":"email|push|both","message":"..."}{"reason":"..."}{"section":"fortress","fields":{"module_name":"..."}}{"settings":{"payment_grace_period_days":"14"}}{"id":"crm","baseline_price":150}Frequently Asked Questions
The most common questions about Aetherius — answered clearly.
Getting started
AI Agent Fortress™
Billing & Payments
Pricing & Modules
Security & Compliance
Troubleshooting
Solutions for the most common deployment and runtime issues.
Deployment issues
| Error | Cause | Fix |
|---|---|---|
| "Page can't be found" on worker URL | workers.dev subdomain not enabled | Dashboard → Workers & Pages → aetherius → Settings → Triggers → Enable workers.dev. Then run wrangler deploy again. |
| Bindings show "not configured" | Must redeploy after adding bindings in dashboard | Run wrangler deploy from the worker folder after adding bindings. |
| "wrangler: command not found" | Node.js needs restart or reinstall | Restart computer after installing Node.js, then try again. |
| "D1_ERROR: no such table" | Schema not applied to database | Run wrangler d1 execute aetherius-db --remote --file=database/schema.sql from the aetherius-v3 folder. |
| "database not found" | Wrong database name in command | The database must be named exactly aetherius-db. Check Cloudflare dashboard → D1. |
API issues
| Symptom | Likely cause | Fix |
|---|---|---|
| 401 Unauthorized on protected routes | Missing or expired token | Include header: Authorization: Bearer your-token. Re-login if token expired. |
| 403 on /api/fortress/* routes | Fortress not enabled for tenant | Run scripts/enable-fortress.sql for your tenant. |
| 400 on POST routes | Missing required fields in JSON body | Check the API reference section for required fields. Include Content-Type: application/json header. |
| 500 Internal Server Error | Database not seeded or misconfigured | Run wrangler tail aetherius in terminal, trigger the request, and read the error message. |
| Bindings show "not configured" at /health | Bindings not added in dashboard, or not redeployed | Add bindings in Cloudflare dashboard Settings → Bindings, then run wrangler deploy. |
View live error logs
To see real-time error messages from your deployed Worker, run:
wrangler tail aetheriusThen trigger the failing request in your browser or Postman. The error will appear in the terminal within seconds.
Glossary
Key terms used throughout the Aetherius platform and documentation.
| Term | Definition |
|---|---|
| Agent | An autonomous AI system that perceives inputs, makes decisions, and takes actions — registered and governed by AI Agent Fortress™ |
| Agent Trust Score™ | Proprietary 0–100 score reflecting an agent's trustworthiness based on behaviour, policy adherence, and risk event history |
| Binding | A connection between a Cloudflare Worker and a Cloudflare service (D1, KV, R2, Workers AI) — configured in the dashboard |
| CMS | Content Management System — the Super-Admin interface for updating all public-facing text without code changes |
| Containment ladder | The 7-level escalation system for AI agent control: Observe → Warn → Restrict → Throttle → Freeze → Quarantine → Terminate |
| D1 | Cloudflare's serverless SQLite database — used by Aetherius for all structured data storage |
| Fortress | Short for AI Agent Fortress™ — the 15-sub-module enterprise AI agent governance platform |
| FIPS 203/204/205 | US federal standards for post-quantum cryptographic algorithms (ML-KEM, ML-DSA, SLH-DSA respectively) |
| Grace period | The window after a payment failure (default: 7 days) during which the user retains access while payment issues are resolved |
| KV | Cloudflare Key-Value store — used by Aetherius for fast edge caching of sessions and settings |
| ML-DSA-65 | Module-Lattice Digital Signature Algorithm — FIPS 204 post-quantum digital signature scheme, used for agent certificates |
| ML-KEM-768 | Module-Lattice Key Encapsulation Mechanism — FIPS 203 post-quantum key exchange scheme |
| Mode 1 / Sovereign AI | AI operation mode where all inference runs on Workers AI edge GPUs — zero data leaves Cloudflare, zero external API keys |
| Ontology | The unified semantic model underlying all Aetherius modules — entities like Account, Employee, and Agent carry relationships and shared context across all 12 clouds |
| R2 | Cloudflare's object storage service — used by Aetherius for document storage, exports, and compliance evidence archives |
| SLH-DSA-128s | Stateless Hash-Based Digital Signature — FIPS 205 post-quantum signature scheme used for long-lived audit evidence |
| Simu-Corp | Aetherius's strategic simulation engine — models 847 business variables across all domains simultaneously in under 3 seconds |
| Super-Admin | The highest-privilege user role in Aetherius — controls billing, RBAC, CMS, pricing, Fortress activation, and all tenant settings |
| Tenant | An enterprise customer organisation within the Aetherius multi-tenant platform |
| Workers AI | Cloudflare's edge AI inference service — enables Mode 1 sovereign AI with no external API keys |
| Wrangler | Cloudflare's command-line tool for deploying Workers, managing D1 databases, and administering Cloudflare services |