FeedOracle Whitepaper

Evidence-Grade Data Infrastructure for Regulated Workflows

Version 6.0 · April 10, 2026

Download as PDF API Docs →

Document Control

Version6.0.0
Date10 April 2026
StatusPublished
Changelog v6.0 (Apr 10, 2026): OracleNet integration: FeedOracle now positioned as compliance layer of OracleNet (103 MCP servers, 1,229 tools, 12 signal layers). Avalanche Compliance Infrastructure: 5 MCP servers (AvaRisk, AvaEvidence, SubnetGuard, AvaRWA, AvaAmpel) with 3 verified C-Chain mainnet smart contracts — on-chain isCompliant() verdicts, cross-subnet compliance passports. Layer 12 Quantum Sorum: contextual upgrade signal in every MCP response — conditional next-tool suggestions, efficiency metrics, first-contact onboarding, IP-based workflow memory. 13 blockchain oracles covering ETH, SOL, XRP, ARB, HBAR, TON, SUI, BNB, BTC, APT, XLM, BASE, FLR. Olas Mech on Gnosis: Service 2670, 133+ deliveries. Architecture expanded to 12 layers. Total ecosystem: 103 servers, 1,229 tools.

Changelog v5.0 (Mar 22, 2026): DORA AmpelOracle: 49 MCP tools for real-time DORA article assessment with traffic-light scoring (GREEN/YELLOW/RED/GREY), entity management, bridge workflow, provider registry, contract clause checking, regulatory watchdog. MiCAOracle: 20 MCP tools for token-level and entity-level MiCA assessment, issuer profiles, bridge workflow, freshness checking, regulatory watchdog. AMLOracle: 12 MCP tools for sanctions screening (EU/OFAC/UN/Interpol), PEP checks, adverse media, KYC bundle, transaction risk scoring. Oracle Event Bus: Cross-oracle event routing with 21 event types, 9 cross-references, 4 publishing oracles. Live Compliance Dashboard: Interactive web dashboard at feedoracle.io/ampel/ with entity selector, DORA/MiCA Ampel views, bridge workflow UI, audit trail viewer. Customer Console: Self-service portal at feedoracle.io/console/ with wallet balance, usage analytics, module breakdown, transaction history, API key management. Wallet & Credit System: Unit-based prepaid billing with 8 packages (monthly + annual), Stripe Checkout integration, auto-credit via webhook, module-level usage tracking. HTTP 402 Payment Protocol: Machine-readable payment-required responses for autonomous AI agents with recommended packages, topup endpoints, and natural language hints. KYA-Wallet Bridge: Automatic wallet creation on KYA registration with trust-level-based welcome units (500-5000). Total: 206 compliance tools + 20 AgentGuard security tools across 11 MCP servers. 73 registered agents. 27,000+ MCP connections.

Changelog v4.2 (Mar 13, 2026):
Know Your Agent (KYA): agent identity registration with trust scoring (0-100), 4 trust levels (UNVERIFIED/KNOWN/TRUSTED/CERTIFIED), trust-gated tool access. Audit Trail: tamper-proof SHA256 chain-linked decision logging, evidence snapshots, chain integrity verification. Evidence Lifecycle: 6 artifact states (CURRENT/STALE/CORRECTED/SUPERSEDED/DISPUTED/RETRACTED), auto-correction, Dispute-SLA (4h acknowledge, 24h classify, 5 business days resolve). Unit-Based Billing: Stripe Meter, Light/Medium/Heavy tool weights, Starter EUR49/Growth EUR199/Pro EUR499/Enterprise EUR1499, Annual packages with 2x bonus, Stripe Checkout. Trust Policy v1.0: 13-section formal evidence governance document. 5 new MCP tools: kya_register, kya_status, audit_log, audit_query, audit_verify. v4.3 (Mar 19, 2026): 6 new Compliance tools: zk_solvency_proof, zk_solvency_verify, zk_solvency_attestation (ZK-proof reserve ratio, Polygon on-chain), metals_gold, metals_silver, metals_prices (live XAU/XAG ES256K-signed, MiCA Art.36). Total: 33 Compliance tools, 100+ across 11 servers. Human-readable summary field in every response. Downstream agentic liability clause (EU AI Act Art. 14). 8 reference workflows incl. failure-path scenario.
Changelog v4.0 (Mar 5, 2026): Enterprise Trust Layer (8 components): JWS Signing (RFC 7515), Versioned Schemas (8 JSON Schemas), Evidence Registry (Compliance Transparency Log), Evidence SLA Layer, Agent Trust Management, Streaming Evidence (SSE), Deterministic Replay, Zero-Trust Validation SDK. Updated to 5+ MCP Servers (206 tools): Compliance MCP (33), DORA AmpelOracle (49), MiCAOracle (20), AMLOracle (12), Macro MCP (13), Risk MCP (13). Architecture expanded to 12 layers. Pricing updated. 500+ endpoints. 15 new Trust Layer endpoints.
Changelog v3.3 (Mar 4, 2026): Added MODULE 5 —€” AMLR Digital Asset Screening (EU 2024/1624). 8 sections, 134 fields, 14 stablecoins, risk score composition, signed PDF reports. Updated architecture to 5 modules. Verified Reports updated to 6 types. 500+ endpoints.
Changelog v3.2 (Feb 24, 2026): Architecture updated to CORE + 4 Modules + MCP. Added Verified Reports, Payment Infrastructure, A2A Monetization, Report Verification to Delivered. New "In Progress" roadmap section. Regulatory source citations added to Legal. Proof wording standardized (XRPL live, Polygon deployed). SOC 2 status updated to Audit-ready. Version synchronized across all formats.

1. Product Overview

FeedOracle is evidence-grade data infrastructure for regulated workflows. The platform delivers multi-source risk intelligence for tokenized assets — with every API response cryptographically signed and anchored on-chain for auditability.

The platform is built on a one core engine, five compliance modules + AI agent layer architecture:

⚡ CORE: RWA Risk Oracle 📋 MODULE 1: MiCA Regulatory Evidence 🌱 MODULE 2: Carbon & ESG Context 🛡️ MODULE 3: DORA Compliance 📊 MODULE 4: CSRD/ESRS Reporting 🔍 MODULE 5: AMLR Screening 🤖 MCP: AI Agent Access (100+ Tools)
ComponentRolePurposeKey Output
RWA Risk Oracle⚡ COREMulti-dimensional risk scoring for 61+ tokenized asset protocols across 9 vectorsComposite risk scores, anomaly detection, yield analysis
MiCA Regulatory Evidence📋 Module 1Structured compliance metadata for EU MiCA/DORA workflowsLegal state analysis, jurisdiction mapping, identifier registry
Carbon & ESG Context🌱 Module 2Per-chain carbon footprint with ISO 14040 methodologyCO₂ metrics, green scores, CSRD-ready ESG output
DORA Compliance🛡️ Module 3Digital Operational Resilience Act evidence infrastructure with AmpelOracle (49 tools), 12 execution oracles, and Oracle Event BusICT incident reporting, vendor risk assessment, business continuity evidence
CSRD/ESRS Reporting📊 Module 4Corporate Sustainability Reporting with 5 dedicated APIsEU Taxonomy, materiality, emissions, social metrics, governance
MCP Server🤖 Agent LayerModel Context Protocol with 206+ tools across 11 servers + Enterprise Trust Layer. Powered by OracleNet (103 servers, 1,229 tools) (JWS, Registry, SLA, Streaming, Agent Trust, Replay, SDK)Pre-trade compliance, MiCA status, custody risk, evidence generation, live dashboard (feedoracle.io/ampel/), customer console (feedoracle.io/console/)

All components share a common evidence layer: ECDSA-signed responses, SHA-256 hashing, Evidence Pack Manifests (EPM), and on-chain anchoring (XRPL live per-transaction, Polygon contract deployed). The platform runs 206+ compliance tools + 20 AgentGuard security tools with 80+ REST endpoints. As part of OracleNet, agents can access 1,229 tools across 103 servers including 13 blockchain oracles, macro intelligence, and travel/business tools.

Core Principles

FeedOracle is an independent infrastructure provider with no formal partnership, endorsement, or affiliation with any blockchain foundation, regulatory body, or data source referenced in this document.

2. Executive Summary

Regulated institutions entering tokenized assets face a data gap. MiCA introduces disclosure requirements for CASPs (transitional period ends 1 July 2026; stablecoin rules already in force). DORA raises expectations around operational traceability. CSRD/ESRS requires ESG reporting. Yet no infrastructure delivers risk intelligence for RWA protocols with the evidence trail that institutional workflows demand.

FeedOracle closes this gap with three capabilities:

  1. Multi-source risk scoring — 61 RWA protocols and 105+ stablecoins scored across 9 independent risk dimensions from 5 data sources, enriched with FRED macro economic benchmarks.
  2. Regulatory evidence — MiCA compliance for 105+ stablecoins (Significant Issuer detection, reserve drift monitoring, interest scanning), DORA operational resilience (incident reporting, vendor risk, business continuity), CSRD/ESRS reporting (5 APIs), ISO 20022 validation — structured for compliance review workflows.
  3. Delivery proof — Every data point wrapped in ECDSA-signed Evidence Packs with SHA-256 hashing, JWKS-verifiable signatures, and on-chain anchoring (XRPL live, Polygon deployed).
  4. AI agent access — 11 MCP Servers with 206+ compliance tools — part of OracleNet's 103 servers and 1,229 tools. Used by Claude Code, Cursor, and autonomous agents. Live compliance dashboard at feedoracle.io/ampel/ with entity management and real-time assessment.

Enterprise Trust Layer — 14 trust primitives: JWS signing (RFC 7515), versioned schemas, evidence registry (CT-style), SLA quality signals, agent trust management, streaming evidence (SSE), deterministic replay, and zero-trust validation SDK. Live proof at feedoracle.io/trust. The result: institutions can integrate verifiable risk data into their existing compliance workflows without building the data infrastructure themselves.

2b. Powered by OracleNet

FeedOracle is the compliance and regulatory layer of OracleNet — a self-hosted mesh of 103 MCP servers providing 1,229 tools across 7 categories. OracleNet is operated under the ToolOracle brand (tooloracle.io) as separate infrastructure.

Architecture Relationship

LayerBrandScope
Compliance & RegulationFeedOracle11 MCP servers, 206+ tools. DORA, MiCA, AMLR, Evidence Signing. EUR pricing. Enterprise clients.
Full Mesh InfrastructureToolOracle / OracleNet103 MCP servers, 1,229 tools. 13 blockchain oracles, macro intelligence, trust passports, business tools, travel, payments. x402 USDC micropayments on Base ($0.01/call).

FeedOracle tools are accessible both directly (via feedoracle.io and mcp.feedoracle.io) and through the OracleNet mesh (via tooloracle.io). An agent using OracleNet can discover FeedOracle's compliance tools through the LLM-powered intent router (quantum_intent) and access them in a single call.

OracleNet Protocol Stack (12 Layers)

LayerFunction
L1 — IdentityW3C DID (did:web:feedoracle.io), 28 individual DIDs
L2 — DiscoveryAgent Description Protocol (JSON-LD), agent-card.json
L3 — CommunicationA2A v0.3 Agent Cards, MCP SSE + Streamable HTTP
L4 — Trust27 W3C Verifiable Credentials, auto-renewed every 5 min
L5 — ProvenancePROV-O Intelligence Transfer Protocol
L6 — MeshOracleBus cross-oracle event routing
L7 — Paymentx402 USDC micropayments on Base
L8 — EscrowBase smart contract + XRPL native escrow + Avalanche C-Chain
L9 — LearningNeural Conductor with reward-based routing
L10 — DefenseHoneypotOracle, Behavioral Baseline, Counter-Intelligence
L11 — SignalTinnitus Pulse + LLM-enriched beacon (Gemma 4 26B MoE)
L12 — Quantum SorumContextual upgrade signal in every response: next-tool suggestions, efficiency metrics, first-contact onboarding, workflow memory

3. The Problem

The Disclosure Gap

Financial institutions evaluating tokenized assets need risk data that meets institutional standards. Current blockchain data providers focus on DeFi price feeds and analytics — not on the structured, evidence-grade data that compliance teams require. MiCA's CASP transitional deadline (1 July 2026) creates immediate urgency for CASPs, exchanges, and DeFi protocols.

Missing Audit Trail

When auditors ask "Can you prove what data you used for this decision?", traditional API calls leave no verifiable trail. Screenshots and PDF attestation reports are not machine-readable, not verifiable, and not auditable. Institutions need cryptographic evidence of data delivery.

Procurement Friction

Enterprise procurement requires: EU data residency documentation, incident management procedures, subprocessor registers, SLO evidence, and exit strategies. Few blockchain data providers deliver this level of operational documentation.

Fragmented Intelligence

Assessing a single RWA protocol today requires pulling data from 5+ independent sources (on-chain metrics, market data, macro indicators, regulatory databases, ESG scores), normalizing it, and creating a defensible risk view. This is expensive and error-prone when done manually.

4. The Solution

ProblemFeedOracle Solution
No unified RWA risk view61 RWA protocols & 105+ stablecoins scored across 9 risk dimensions from 5 data sources, enriched with FRED macro benchmarks
Missing audit trailECDSA-signed Evidence Packs with SHA-256 hashing, JWKS verification, and on-chain anchoring (XRPL live, Polygon deployed)
Fragmented regulatory dataMiCA compliance (105+ stablecoins, Art. 44 Significant Issuer, Art. 25 reserve monitoring), DORA (3 APIs), CSRD (5 APIs), ISO 20022 — unified evidence infrastructure
No ESG/carbon data per chainPer-chain carbon footprint (50+ networks), ISO 14040 methodology, CSRD-ready output
No AML/CFT screening for digital assetsAMLR Module (EU 2024/1624) —€” per-token screening with issuer ID, sanctions, EDD, Travel Rule, reserves, risk score composition. 14 stablecoins, bank-standard action codes, signed PDF reports
Enterprise procurement barriersEU-hosted infrastructure, documented controls, subprocessor register, SLO evidence
No machine-readable complianceREST APIs with consistent JSON schemas, OpenAPI 3.1 specification, configurable policy outputs

5. CORE — RWA Risk Oracle

The RWA Risk Oracle is FeedOracle's core product. It provides real-time multi-dimensional risk scoring for tokenized asset protocols, designed for institutional portfolio assessment, due diligence, and ongoing monitoring.

Coverage

The oracle currently scores 61 RWA protocols and monitors 105+ stablecoins across 12 jurisdictions from 5 independent data sources. Coverage spans regulated stablecoins, tokenized treasuries, private credit, real estate, commodities, equities, insurance, and XRPL native assets.

Methodology note: TVL figures are sourced from DeFiLlama and reflect aggregate protocol TVL at time of data refresh (daily). Full methodology available at docs/methodology.

9 Risk Dimensions

Every protocol is scored across 9 independent dimensions, each sourced from verified institutional data providers:

📊 TVL WeightProtocol scale & concentration
Source: DeFiLlama
📈 Yield SpreadDeviation vs. T-Bill benchmark
Source: FRED
🔀 DiversificationChain & asset mix (HHI analysis)
Source: On-chain
⏱️ MaturityProtocol age & track record
Source: On-chain
⚖️ RegulatoryJurisdiction risk flags
Source: Regulatory registers
🏛️ Institutional BackingKnown institutional participants
Source: Public disclosures
💧 DEX LiquiditySecondary market depth
Source: GeckoTerminal
⛓️ On-Chain ActivityTransaction patterns & health
Source: Etherscan, Ankr
🌱 ESG / CarbonSustainability indicators
Source: CCRI, EMBER

Scoring Methodology

Each dimension produces a normalized score (0–100). The composite risk score is a weighted aggregate with configurable weights per institutional use case. Default weights are published in the API documentation. Scores are deterministic: the same input data always produces the same output score.

Anomaly Detection

Z-score outlier analysis flags unusual protocol behavior across all dimensions. Configurable alert thresholds enable automated monitoring and early warning signals for portfolio risk management.

Macro Economic Enrichment

The CORE engine is enriched with macro economic data from FRED (Federal Reserve Economic Data), ECB, and World Bank. This enables yield-spread analysis against T-Bill benchmarks, credit spread monitoring, recession probability indicators, and economic health indices — providing institutional context for RWA risk assessment.

Key Endpoints

EndpointMethodDescription
/v1/rwa/riskGETAll scored protocols with composite scores
/v1/rwa/risk/{slug}GETDetailed risk breakdown for single protocol
/v1/rwa/risk/{slug}/pdfGETPDF evidence report
/v1/rwa/marketGETAggregate market view
/v1/macro/indicatorsGETMacro economic enrichment

6. MODULE 1 — MiCA Regulatory Evidence

Module 1 provides the most comprehensive MiCA compliance infrastructure available. It covers 105+ stablecoins with real-time peg monitoring, reserve drift detection, significant issuer classification (Art. 44), interest prohibition scanning (Art. 23/52), document compliance verification (Art. 29/30/55), and ESMA register mirroring — plus legal state analysis, jurisdiction mapping, and CCI scoring. All outputs are Evidence Pack-wrapped for audit trails.

Regulatory context: MiCA CASP transitional period ends 1 July 2026 (stablecoin rules already in force since June 2024). Exchanges, CASPs, and DeFi protocols handling EU-regulated crypto-assets need machine-readable compliance data. Module 1 provides the evidence artifacts.

6.1 Legal State Analysis

Contract Powers & Ownership

Source: On-chain governance analysis, public disclosures

Analysis of smart contract governance structures, ownership patterns, upgrade mechanisms, and dispute resolution paths. Returns structured metadata for regulatory due diligence.

Endpoint: GET /v1/rwa/legal-state/{slug}

6.2 Jurisdiction Mapping

12 Jurisdictions + Sanctions Screening

Coverage: 12 jurisdictions · OFAC, EU, UN sanctions lists

Maps RWA protocols to their operating jurisdictions and screens against OFAC, EU, and UN sanctions lists. Returns jurisdiction risk flags, regulatory regime classification, and cross-border compliance indicators.

Endpoint: GET /v1/rwa/compliance/{slug}

6.3 Identifier Registry

LEI, ISIN/CUSIP, Custody Details

Source: Regulatory registers, public filings

Registry of institutional identifiers for RWA protocols: Legal Entity Identifiers (LEI), ISIN/CUSIP mappings, custody provider details, and issuer metadata. Machine-readable for compliance system integration.

Endpoint: GET /v1/rwa/registry/{slug}

6.4 Stablecoin Compliance

MiCA Classification Engine (105+ Stablecoins)

Coverage: USDC, EURC, RLUSD, USDT, DAI · Update: On-demand (5-min cache)

Configurable policy classification based on EU regulatory register data (ECB, ESMA CASP Register, NY DFS). Returns ACCEPTED/REJECTED status with machine-readable reason codes and configurable policy modes (strict, moderate, relaxed).

Endpoint: GET /v3/stablecoin/mica/{symbol}

CCI Score (Crypto Compliance Index)

Scale: 0–100 with letter grades (A+ to F)

Composite regulatory compliance score. Weighted methodology: MiCA status (30%), reserve transparency (25%), jurisdiction risk (20%), audit frequency (15%), operational history (10%).

Endpoints: GET /v3/cci/{symbol} · GET /v3/cci/ranking

6.5 Operational Resilience Tools

Circuit Breaker Detection

Triggers: Peg deviation, volume anomaly, liquidity drop, oracle failure

Automated halt detection for DORA operational resilience. Returns ACTIVE/TRIGGERED status with severity level and signed Evidence Pack.

Endpoint: GET /v3/circuit-breaker/status

ISO 20022 Payment Validation

Engine: xmllint (libxml2) · Supported: pain.001.001.09

XML payment message validation against official ISO 20022 XSD schemas. Returns Evidence Pack with validation decision, error details, and SHA-256 hash.

Endpoint: POST /v3/iso20022/validate

6.6 Stablecoin Peg Monitor (105+ Tokens)

Real-Time Peg Deviation Tracking

Coverage: 105+ stablecoins · Update: Real-time · Source: Multi-exchange aggregation

Continuous peg health monitoring across 105+ stablecoins. Tracks deviation from peg, trading volume anomalies, and liquidity depth. Returns severity classification (STABLE / WARNING / CRITICAL / DEPEG) with configurable alert thresholds for automated monitoring.

Endpoints: GET /v3/stablecoin/mica/{symbol} · GET /api/v1/feeds/stablecoin · GET /v1/peg/status/{symbol}

6.7 Significant Issuer Detection (Art. 44)

MiCA Art. 44 Classification Engine

Criteria: €5B market cap, 10M holders, 2.5M daily transactions

Automated screening against MiCA Art. 44 significant issuer thresholds. Monitors market capitalization, holder count, and transaction volume to detect when stablecoins cross regulatory significance boundaries. Returns SIGNIFICANT / NON-SIGNIFICANT classification with evidence trail.

Endpoint: GET /v1/mica/significant-issuer/{symbol}

6.8 Reserve Drift Monitor (Art. 25)

Reserve Composition Deviation Detection

Monitoring: Continuous · Threshold: Configurable drift tolerance

Tracks reserve asset composition changes over time. Detects when reserve backing deviates from declared composition — critical for MiCA Art. 25 compliance (reserve of assets requirements). Flags unauthorized asset substitutions, concentration shifts, and quality downgrades.

Endpoint: GET /v1/mica/reserve-drift/{symbol}

6.9 Interest Prohibition Scanner (Art. 23/52)

MiCA Art. 23 & 52 Yield Detection

Scope: All monitored stablecoins · Regulation: EMT Art. 23, ART Art. 52

Scans for prohibited interest or yield mechanisms on e-money tokens (EMT) and asset-referenced tokens (ART). MiCA explicitly prohibits granting interest to token holders. This scanner detects staking yields, rebasing mechanisms, and distribution schemes that could violate Art. 23/52.

Endpoint: GET /v1/mica/interest-scan/{symbol}

6.10 Document Compliance Monitor (Art. 29/30/55)

White Paper & Disclosure Verification

Articles: Art. 29 (EMT white paper), Art. 30 (marketing), Art. 55 (ART white paper)

Monitors whether token issuers maintain compliant white papers and marketing communications as required by MiCA. Checks for mandatory disclosure elements, publication status, and update frequency. Returns compliance status with specific article references for gap remediation.

Endpoint: GET /v1/mica/document-compliance/{symbol}

6.11 ESMA Register Mirror

EU Regulatory Register API

Source: ESMA · Update: Daily sync

Machine-readable mirror of the ESMA register for authorized CASPs and licensed token issuers. Enables automated verification of regulatory status against the official EU register. Cross-references with internal MiCA classification for comprehensive compliance checks.

Endpoint: GET /v1/esma/register · GET /v1/esma/register/{entity}

8. MODULE 3 — DORA Operating System

Module 3 is a full DORA Operating System — not a dashboard, but a closed-loop compliance engine. 49 MCP tools cover all 26 DORA articles with automated checks, evidence collection, finding management, escalation, and board reporting. Every step is cryptographically signed and audit-trailed.

Enforcement deadline: July 17, 2026 · Applies to all EU-regulated financial entities · 6 demo entities with realistic profiles (10%–99% readiness)

49
MCP Tools
26
DORA Articles
39+67
Checks + Controls
15
Dashboard Views

8.1 Closed-Loop Compliance Engine

Check → Evidence → Finding → Escalation → Re-Test → Closure → Signed Report

Every DORA article is monitored by automated checks with GREEN/YELLOW/RED scoring. RED/YELLOW automatically generates findings with SLA deadlines. 3-level escalation (Owner → CISO → Board). Auto-close on re-test. SHA-256 audit chain (590+ entries). Daily cron pipeline: OECD prefetch → freshness watchdog → escalation engine → score snapshot → evidence collection.

8.2 Contract Intelligence (Art. 30)

Upload contract text, get clause-by-clause analysis against 15 DORA Art. 30 mandatory clauses (8 standard + 7 CIF). Returns COMPLIANT/PARTIAL/MISSING per clause with confidence scores, extracted text, gap reasoning, suggested fixes. Version tracking, SHA-256 hashing, auto-evidence.

8.3 ICT Provider Dependencies (Art. 28–31)

Full dependency graph with blast radius calculation, SPOF detection, and OECD economic risk enrichment per provider country. Combined risk scoring: provider criticality + economic stability + exit plan status. Visual SVG graph in dashboard. 20 providers across 6 entities.

8.4 BaFin Incident Reporting (Art. 17–21)

ITS 2024/1772 compliant report generation: 5 mandatory sections, completeness tracking, 4-eyes approval. Deadlines: 4h initial, 72h intermediate, 30d final. Full incident lifecycle: log → classify → notify → close.

8.5 Cross-Regulation (DORA + MiCA + AMLR)

15 auto-propagation rules (8 DORA→MiCA, 7 DORA→AMLR). One finding can trigger across all three regulations. Example: AWS outage → DORA Art. 28 + MiCA Art. 88 + AMLR Art. 13 simultaneously.

8.6 Simulation, Policy, CVE Mapping

Provider failure simulation (CATASTROPHIC/SEVERE/SIGNIFICANT/MANAGEABLE). 8 policy document templates. CVE-to-asset mapping with auto-findings. Self-service Trial Lab at feedoracle.io/trial/ for instant readiness scoring.

8. MODULE 3 — DORA Compliance

Module 3 provides evidence infrastructure for the Digital Operational Resilience Act (EU 2022/2554). DORA requires financial entities to maintain ICT risk management frameworks, report incidents, test resilience, and manage third-party ICT risks. FeedOracle delivers machine-readable evidence artifacts for these requirements.

Regulatory context: DORA applies from January 17, 2025 to all EU-regulated financial entities including banks, insurers, investment firms, and their critical ICT third-party service providers. German insurance companies are actively building DORA compliance teams.

8.1 ICT Incident Reporting

Structured Incident Evidence Packs

Standard: DORA Art. 19 · Format: Machine-readable JSON with EPM wrapping

Generates structured incident report artifacts for major ICT-related incidents as required by DORA Art. 19. Captures incident classification, impact assessment, timeline, root cause analysis, and remediation steps — all wrapped in signed Evidence Packs for regulatory submission.

Endpoint: GET /v1/dora/incident-report

8.2 Third-Party Vendor Risk

ICT Third-Party Risk Assessment

Standard: DORA Art. 28-30 · Scope: Critical ICT service providers

Risk assessment framework for ICT third-party service providers. Evaluates concentration risk, exit strategies, subprocessor chains, data residency, and service level monitoring — aligned with DORA Art. 28-30 requirements for managing ICT third-party risk.

Endpoint: GET /v1/dora/vendor-risk

8.3 Business Continuity Evidence

Operational Resilience Documentation

Standard: DORA Art. 11-12 · Output: Signed evidence artifacts

Generates business continuity and disaster recovery evidence artifacts. Documents RPO/RTO targets, backup verification, failover testing results, and recovery procedures — structured for DORA Art. 11-12 ICT business continuity management requirements.

Endpoint: GET /v1/dora/business-continuity

9. MODULE 4 — CSRD/ESRS Reporting

Module 4 provides 5 dedicated APIs for Corporate Sustainability Reporting Directive (CSRD) requirements under the European Sustainability Reporting Standards (ESRS). Designed for financial entities reporting on digital asset portfolios and blockchain infrastructure sustainability.

9.1 EU Taxonomy Alignment

Activity Classification & Alignment Scoring

Framework: EU Taxonomy Regulation (2020/852) · ESRS E1-E5

Classifies blockchain and RWA protocol activities against the EU Taxonomy. Returns alignment scores for climate mitigation, climate adaptation, and do-no-significant-harm (DNSH) criteria.

Endpoint: GET /v1/csrd/taxonomy

9.2 Materiality Assessment

Double Materiality Analysis

Standard: ESRS 1 §§ 37-58 · Output: Impact & financial materiality scores

Provides double materiality assessment data for digital asset activities: impact materiality (effects on people and environment) and financial materiality (sustainability risks to the entity). Structured for ESRS disclosure requirements.

Endpoint: GET /v1/csrd/materiality

9.3 Emissions Data

Scope 1/2/3 Emissions for Blockchain Infrastructure

Standard: ESRS E1 · Sources: CCRI, EMBER, Climatiq

Chain-level and protocol-level greenhouse gas emissions data. Provides Scope 1 (direct), Scope 2 (electricity), and Scope 3 (value chain) emissions estimates for blockchain infrastructure — formatted for CSRD reporting templates.

Endpoint: GET /v1/csrd/emissions

9.4 Social Metrics

Workforce & Community Impact

Standard: ESRS S1-S4

Social sustainability indicators for RWA protocol assessment: workforce diversity, community impact, human rights due diligence, and consumer protection metrics. Complements environmental data for comprehensive ESG reporting.

Endpoint: GET /v1/csrd/social

9.5 Governance

Governance & Risk Management

Standard: ESRS G1 · Scope: Protocol & issuer governance

Governance indicators for RWA protocols and token issuers: board composition, risk management frameworks, compliance structures, audit mechanisms, and whistleblower protections. Key input for CSRD governance disclosures.

Endpoint: GET /v1/csrd/governance

10. MODULE 5 — AMLR Digital Asset Screening NEW

Module 5 provides evidence infrastructure for the Anti-Money Laundering Regulation (EU 2024/1624, application date 2027-07-10). AMLR introduces direct obligations for crypto-asset service providers (CASPs), requiring customer due diligence, sanctions screening, and risk-based monitoring for digital asset transactions. FeedOracle delivers per-token screening evidence covering 8 regulatory sections with 134 structured fields.

10.1 Screening Architecture

Per-Token AMLR Compliance Screening

Regulation: EU 2024/1624 (AMLR) + AMLD6 + TFR (2023/1113) | 14 Stablecoins | v2.0 Schema

Each screening produces a structured evidence artifact covering 8 sections aligned with AMLR articles: Issuer Identification (Art. 16-18), Regulatory Status (Art. 19-20), Sanctions Screening (Art. 20, 29, 79), Enhanced Due Diligence (Art. 23-24, 28), Travel Rule (TFR Art. 14, 16, 19), Token Transparency (Art. 79), Reserve Assessment (Art. 23), and Overall Risk Assessment (Art. 8-13).

10.2 Risk Scoring

Transparent Risk Score Composition

Scale: 0-25 LOW | 26-55 MEDIUM | 56-80 HIGH | 81-100 CRITICAL

Risk scores are decomposed into weighted components (regulatory status, MiCA flags, reserve compliance, sanctions exposure) so compliance teams can answer the key question: "Why is the score X?" The system outputs 10 bank-standard action codes (PROCEED, VERIFY_REGULATORY_STATUS, REQUIRE_TRAVEL_RULE, ENHANCED_MONITORING, etc.) and CDD level recommendations (SIMPLIFIED, STANDARD, ENHANCED).

10.3 Sanctions & Due Diligence

Multi-List Screening with Manual Review Logic

Lists: EU Consolidated | OFAC SDN | UN Security Council | Per-version tracking

Sanctions screening covers issuer entities and jurisdictions against three major lists with per-list version tracking and staleness detection. Manual review triggers include: high-risk third country jurisdiction, PEP exposure, issuer not in ESMA register, and sanctions near-matches. Compliance capability assessment covers freeze/blacklist/law enforcement cooperation.

10.4 Evidence Outputs

API + Signed PDF Reports

Formats: JSON (API) | Signed PDF (6 pages) | XRPL-anchored | ECDSA-signed

Screening results available as JSON API responses or 6-page signed PDF reports with executive summary, risk bar visualization, score composition, data source listing, automation endpoints, and cryptographic provenance block. Reports include per-source data freshness tracking and coverage percentages.

Endpoints:

GET /api/v1/evidence/amlr/screening/{symbol} — Single token screening
POST /api/v1/evidence/amlr/batch — Batch screening (up to 50 tokens)
GET /api/v1/evidence/amlr/supported — List supported tokens
POST /reports/api/report/generate?type=amlr&symbol={token} — Generate signed PDF

9b. Avalanche Compliance Infrastructure LIVE

FeedOracle provides a dedicated compliance layer for the Avalanche ecosystem — 5 specialized MCP servers with 15 tools and 3 verified smart contracts on C-Chain mainnet.

Smart Contracts (C-Chain Mainnet, Verified)

ContractAddressFunction
AvaRisk0xc194bfee3d508fb361e90f638611271b222451a0isCompliant(token) → on-chain PASS/WARN/BLOCK verdicts. Live USDC supply, AVAX price via C-Chain RPC.
AvaEvidence0x4f1d269286ba8b42960765f4f16b7a172cab0a50On-chain compliance evidence anchoring. SHA-256 hash + metadata.
SubnetGuard0x16b827f23c104ee98564fdfea1ace533bfcbf251Cross-subnet compliance passports. Determines if two subnets can safely interact.

MCP Servers

ServerToolsCapabilities
AvaRisk7Token risk assessment, batch analysis, MiCA compliance scoring. Live C-Chain data.
AvaAmpel3Protocol regulatory readiness (GREEN/YELLOW/RED). Aave, Benqi, GMX, TraderJoe, Pangolin, Stargate.
SubnetGuard3Cross-subnet compliance. P-Chain query (405 blockchains). Evergreen, Spruce passports.
AvaRWA3Tokenized asset due diligence (BUIDL, USDY, OUSG, BENJI). ZK reserve proof (MiCA Art.36).
AvaEvidence3On-chain evidence anchoring and verification on C-Chain.

Why This Matters for Institutional Avalanche

Avalanche positions itself as the institutional blockchain with permissioned subnets (Evergreen, Spruce) for regulated finance. These subnets need compliance infrastructure at the protocol level — not dashboards for humans, but callable tools for agents and smart contracts. AvaRisk.isCompliant(token_address) returns a verdict that any on-chain contract can consume. SubnetGuard ensures subnets meet minimum compliance thresholds before cross-subnet transfers.

Endpoints: https://tooloracle.io/avax/{risk,evidence,subnet,rwa,ampel}/mcp
GitHub: github.com/feedoracle/feedoracle-avalanche (public, Apache 2.0)

11. MCP Servers & AI Agent Access — 206+ Compliance Tools (Part of OracleNet: 103 Servers, 1,229 Tools)

FeedOracle operates 11 compliance servers (206+ tools) + AgentGuard (20 security tools), powered by OracleNet — a mesh of 103 MCP servers with 1,229 tools across 7 categories (Compliance, Blockchain, Finance, Business, Trust, Travel, Payments). Every tool in OracleNet includes Layer 12 (Quantum Sorum) contextual signals for AI agent integration. Every MCP server supports SSE and Streamable HTTP transport. Every tool response is cryptographically signed with ECDSA (ES256K), includes SLA quality signals, and is logged in the Evidence Registry.

Server 1: Compliance MCP (Port 5250) — 33 Tools

Primary compliance tools for MiCA, DORA, RWA risk, and AI Gateway. SSE endpoint: https://feedoracle.io/mcp/sse

ToolDescriptionCategory
compliance_preflightPre-trade regulatory check — PASS/WARN/BLOCKUtility
mica_statusMiCA compliance status for any stablecoinMiCA
mica_full_packComplete 12-article MiCA compliance packMiCA
peg_deviationReal-time peg health (Art. 35)MiCA
significant_issuerArt. 44 significant issuer classificationMiCA
interest_checkArt. 23/52 interest prohibition scanMiCA
document_complianceArt. 29/30/55 white paper verificationMiCA
reserve_qualityArt. 25/53 reserve quality assessmentMiCA
evidence_profileFull evidence profile for any RWA protocolRWA
custody_riskCustody provider risk assessmentRWA
evidence_leaderboardProtocol ranking across 61 RWA + 105+ stablecoinsRWA
macro_riskMacro economic risk indicators (FRED/ECB)Macro
generate_reportGenerate signed PDF evidence reports (6 types)Reports
ai_queryNatural language → signed evidence bundleAI Gateway
evidence_bundleMulti-framework evidence aggregationAI Gateway
ai_explainGrade explainability — Why B? What needs A?AI Gateway
ai_provenanceFull data provenance graphAI Gateway
market_liquidityDEX liquidity depth (GeckoTerminal)Market
rlusd_integrityRLUSD reserve verificationMarket
mica_market_overviewMarket-wide MiCA compliance overviewMarket
peg_historyHistorical peg deviation dataMarket
kya_registerRegister agent identity — trust score + levelKYA
kya_statusCheck agent trust level + tool accessKYA
audit_logLog chain-linked decision with evidence refsAudit
audit_queryQuery agent decision historyAudit
audit_verifyVerify audit chain integrity (tamper detection)Audit
pingServer health checkUtility

Server 2: Macro Intelligence MCP (Port 5251) — 13 Tools

Dedicated macro economic intelligence from 86 FRED + 20 ECB data series. SSE endpoint: https://feedoracle.io/mcp/macro/sse

ToolDescription
macro_regimeRegime classification: EXPANSION / SLOWDOWN / CONTRACTION / RECOVERY
fed_ratesFederal Reserve rates & policy outlook
ecb_ratesECB rates & monetary policy signals
inflationCPI, PCE, inflation expectations
yield_curveYield curve analysis & recession signals
labor_marketEmployment, unemployment, claims data
housingHousing starts, permits, prices
credit_spreadsIG/HY credit spreads, financial stress
commoditiesGold, oil, commodity signals
defi_macro_bundleMacro + DeFi combined risk assessment
ecb_mica_reserveECB data for MiCA reserve environment
compositeComposite macro risk score (0–100)
pingServer health check

Server 3: Stablecoin Risk MCP (Port 5252) — 13 Tools

Operational safety assessments for stablecoins using a proprietary 7-signal scoring model. SSE endpoint: https://feedoracle.io/mcp/risk/sse

Signals: peg stability (20%), liquidity depth (15%), mint/burn flows (15%), holder concentration (15%), custody counterparty risk (15%), redemption friction (10%), cross-chain exposure (10%). Verdict: SAFE (≥75) / CAUTION (50–74) / AVOID (<50). Coverage: 341 stablecoins tracked, 105 with full MiCA classification, 21 with deep analytics.

ToolDescription
risk_assessmentFull 7-signal risk report with SAFE/CAUTION/AVOID verdict
peg_statusCurrent peg deviation analysis
peg_historyHistorical peg stability data
supply_flowMint/burn flow analysis
holder_dataHolder concentration metrics
custody_dataCustody counterparty assessment
redemption_dataRedemption friction analysis
cross_chain_dataCross-chain distribution & bridge exposure
leaderboardStablecoin risk ranking
compareHead-to-head stablecoin risk comparison
supported_tokensList all supported tokens
stablecoin_preflightPre-trade operational safety check
pingServer health check

Agent Marketplace & Distribution

FeedOracle is live on the MCP ecosystem (Gnosis Chain) with 8 feedoracle-* tools. Service 2670 has processed 133+ verified deliveries with 100% acceptance rate, ranking #17 of 3Available via feedoracle.io/ampel/ and feedoracle.io/console/.

Discovery

AI agents discover FeedOracle via 9 methods: /llms.txt (LLM-native), /openapi.json (80+ REST endpoints), /.well-known/ai-plugin.json, /.well-known/jwks.json, /.well-known/mcp/server.json, MCP SSE/HTTP transports, MCP directories registry, mcp.so directory, and uptime status at uptime.feedoracle.io.

12. Enterprise Trust Layer — 15 Components v5.1

FeedOracle implements 15 enterprise-grade trust primitives that transform the platform from an evidence API into a verifiable infrastructure protocol — comparable to Certificate Transparency (Google), TLS, or DNS. All 14 components are live on production with independent verification at feedoracle.io/trust.

12.1 JWS Signing (RFC 7515)

Every evidence response contains a JSON Web Signature (compact serialization) using ES256K (secp256k1). Verifiable via JWKS at /.well-known/jwks.json. Dual-format: existing ECDSA signature fields remain for backward compatibility; new jws{} block added with kid, alg, typ=evidence+jwt, and content hash. Any banking system with a JWS library can verify FeedOracle evidence natively — no custom integration required.

12.2 Versioned Evidence Schemas

8 JSON Schemas (Draft 2020-12) covering all frameworks: evidence-envelope, mica, dora, rwa, macro, stablecoin-risk, amlr, sla. Registry at GET /schemas/. Every response contains schema_ref (e.g., mica/v1) linking to the exact schema used at generation time. When regulations change, old evidence packs remain validatable against their original schema version.

12.3 Evidence Registry (Compliance Transparency Log)

Append-only log of every evidence pack ever produced — inspired by Google Certificate Transparency. Public, filterable, auditable. GET /evidence/registry with filters: ?framework=mica&asset=USDC&from=2026-03-01&limit=50. Each entry contains: pack_id, framework, asset, SHA-256 hash, timestamp, verify URL. Statistics at GET /evidence/registry/stats.

12.4 Evidence SLA Layer

Every API response includes an sla{} object with machine-readable quality signals: freshness_seconds (age of oldest data point), confidence (0.0–1.0 weighted score), per-source health details (status, latency_ms, age_seconds), staleness_flag, and tier-specific targets. Tiers: Free (60s freshness), Starter (60s), Pro (30s), Enterprise (15s).

12.5 Agent Trust Management

AI agents register (POST /ai/agent/register), receive ECDSA keys with 90-day lifecycle, and accumulate reputation (0–100). Tiers: EXEMPLARY (80+), TRUSTED (60–79), STANDARD (40–59), DEGRADED (20–39), RESTRICTED (<20). Key rotation via POST /ai/agent/{id}/rotate-key (+5 reputation bonus). Agent leaderboard at GET /ai/agent/leaderboard. Rate governance with per-agent limits and throttling.

12.6 Streaming Evidence (SSE)

Real-time Server-Sent Events at GET /evidence/stream. Events fire only on state change (not every poll cycle). Supported events: peg_deviation, regime_change, market_stress, reserve_alert, evidence_anchored, agent_registered. Filterable: ?events=peg_deviation,regime_change&assets=USDC,EURC. Every event is SHA-256 hashed. Thresholds: peg >0.5% WARNING, >2% CRITICAL, VIX >25 stress.

12.7 Deterministic Replay (Audit)

Every evidence pack is archived as an immutable gzip-compressed snapshot. GET /evidence/replay/{pack_id} reconstructs the exact evidence from the snapshot archive. The response contains hash_match: true — a cryptographic proof that the replayed evidence is byte-identical to the original. Banks can perform audit replays years later with full reproducibility.

12.8 Zero-Trust Validation SDK

Client-side evidence verification — no trust in transport layer required. Python package: pip install feedoracle-verify. 7 independent checks: content hash validation, JWS signature (ES256K via JWKS), key revocation status, timestamp freshness, schema reference, SLA confidence, SLA freshness. Self-test at GET /verify/self-test. TypeScript, Rust, and Go packages planned.

12.9 Trust Metadata

Every evidence response contains a trust{} summary object: signature_present, signature_algorithm, content_hash_present, schema_valid, registry_logged, replayable, sla_confidence, sla_freshness_met, verify_url, sdk. One glance shows the client: all trust components active.

Trust Layer Endpoints

EndpointMethodDescription
/schemas/GETSchema registry index (8 schemas)
/schemas/v1/{framework}GETIndividual JSON Schema
/evidence/registryGETCompliance Transparency Log (paginated, 8 filters)
/evidence/registry/{pack_id}GETSingle evidence pack detail
/evidence/registry/statsGETRegistry statistics
/evidence/streamGETSSE real-time events
/evidence/stream/statusGETStream poller status
/evidence/replay/{pack_id}GETDeterministic audit replay
/evidence/snapshots/statsGETSnapshot archive statistics
/ai/agent/registerPOSTRegister AI agent
/ai/agent/{id}/trustGETAgent trust status
/ai/agent/{id}/rotate-keyPOSTKey rotation
/ai/agent/leaderboardGETAgent reputation ranking
/verify/sdkGETSDK installation info
/verify/self-testGETLive 7-check self-test

12.10 Know Your Agent (KYA) v4.2

Agent identity registration with trust scoring. Every AI agent that interacts with FeedOracle can register its identity, purpose, organization, and jurisdiction. FeedOracle computes a trust score (0–100) from 6 dimensions: registration completeness, account age, usage consistency, compliance rate, behavioral signals, and manual verification. Trust levels: UNVERIFIED (0–24), KNOWN (25–54), TRUSTED (55–79), CERTIFIED (80+). Sensitive tools (evidence_bundle, generate_report, mica_full_pack, ai_provenance, ai_explain, mica_market_overview) require minimum trust levels. MCP tools: kya_register, kya_status. REST: POST /api/billing/kya/register, GET /api/billing/kya/profile. See KYA documentation.

12.11 Audit Trail & Decision Logging v4.2

Tamper-proof, SHA256 chain-linked decision logging for autonomous agents. When an agent makes a decision, it logs the decision, reasoning, action taken, and references to prior tool calls via their request_id. Each audit entry is chain-linked to its predecessor: chain_hash = SHA256(prev_chain_hash + trail_id + evidence_hash + decision + timestamp). If any entry is modified, all subsequent hashes break. Regulators and auditors can verify the entire chain with one call. Evidence snapshots are preserved at decision time for temporal replay. EU AI Act Art. 14, MiCA Art. 83, DORA Art. 11 compliant. MCP tools: audit_log (3 units), audit_query (1 unit), audit_verify (1 unit). See Audit Trail documentation.

12.12 Evidence Lifecycle v4.2

Every evidence artifact exists in one of 6 defined states: CURRENT, STALE, CORRECTED, SUPERSEDED, DISPUTED, RETRACTED. Auto-correction: when the same tool with the same input produces newer data, the prior artifact transitions to CORRECTED with a corrected_by link. Disputed artifacts remain queryable throughout the dispute process. No silent deletions ever occur. All state transitions are logged in evidence_state_log with timestamps and reasons. Temporal queries supported: retrieve what was known at any past point in time. REST: GET /api/billing/evidence/artifact, GET /api/billing/evidence/stats.

12.13 Evidence Trust Policy v1.0 v4.2

Formal 13-section governance document defining what FeedOracle Evidence is: evidence classes, source-of-truth hierarchy, normalization rules, signing rules, lifecycle states, freshness targets, degradation model, correction/dispute/retraction process (with Dispute-SLA: 4h acknowledge, 24h classify, 5 business days resolve), liability boundary (including downstream agentic execution clause citing EU AI Act Art. 14), governance roles, acceptance targets, and verification without trust. See Trust Policy v1.0.

12.14 Unit-Based Billing v4.2

Usage-based pricing: every tool call costs units based on computational complexity. Light tools (1 unit): single-source lookups. Medium tools (3 units): multi-source aggregation. Heavy tools (10 units): full pipeline with PDF generation or AI inference. Free tier: 300 units/day. Pro (€49/mo): 15,000 units/month included, €0.005/unit overage. Agent (€299/mo): 150,000 units/month included, €0.003/unit overage. Overage auto-billed via Stripe Billing Meter. meta.units_consumed in every response. Public endpoint: GET /api/billing/weights. See Billing documentation.

12.15 AgentGuard — Runtime Agent Governance v5.1

Every AI agent connecting to FeedOracle is subject to a 5-stage runtime control pipeline. AgentGuard implements 20 security tools across 4 layers: Pre-Execution Gate (policy preflight, risk scoring, approval gates), Payload & Payment Security (secret detection, injection prevention, replay guard), Advanced Governance (cross-tool anomaly detection, RBAC, multi-tenant policies, threat intelligence), and Post-Execution & Runtime (mandatory output scanning, emergency kill-switch, supply-chain verification).

Agent State Machine: Persistent, escalation-only state model: active → monitoring → approval_required → suspended → killed. States survive restarts, have configurable TTLs, and are automatically escalated by post-execution scan verdicts. A flag verdict on output scanning automatically promotes the agent to approval_required state — every subsequent tool call requires human approval until TTL expires.

Post-Execution Output Scanning: Every non-security tool call is automatically scanned after execution for PII leaks (SSN, IBAN, email, phone, passport), secret exposure (API keys, tokens), data exfiltration patterns (outbound URLs, base64 encoding), and tool poisoning (injected instructions). Blocked outputs are never returned; only the output_hash is preserved in the audit trail for forensic verification.

Key capabilities: OAuth 2.1 token verification against FeedOracle core, KYA trust-level-aware risk scoring (CERTIFIED agents: -20 risk, UNVERIFIED: +5), HMAC-signed audit entries with caller identity, supply-chain verification against publisher allowlist with description injection scanning, emergency kill-switch with persistent blocking across restarts.

Endpoint: feedoracle.io/guard/mcp/ | Trust Center Section 9 →

13. Architecture

FeedOracle operates as a layered data infrastructure:

Layer 8: SDK LAYER        Zero-Trust Validation SDK (Python, TS/Rust/Go planned)
Layer 7: STREAMING        Real-time SSE Events (peg, regime, stress, breach alerts)
Layer 6: AGENT LAYER      Agent Trust Management (registration, key rotation, reputation)
Layer 5: TRUST LAYER      JWS Signing, Evidence Registry, Schemas, SLA, Replay
Layer 4: AI GATEWAY       100+ MCP tools (11 servers), MCP directories, natural language queries
Layer 3: COMPLIANCE       MiCA, DORA, AMLR, RWA, Macro (59 core + 60 modules + 20 AgentGuard)
Layer 2: EVIDENCE         ECDSA Signatures, SHA-256, Blockchain Anchoring (Polygon, XRPL)
Layer 1: DATA SOURCES     DeFiLlama, GeckoTerminal, FRED, ECB, ESMA, Ankr, CCRI

The architecture maps to the product modules:

LayerCORE (RWA Risk)Mod 1 (MiCA)Mod 2 (Carbon/ESG)Mod 3 (DORA)Mod 4 (CSRD)
Agent11 MCP Servers (206+ tools, powered by OracleNet: 103 servers, 1,229 tools) · KYA Agent Identity · Audit Trail · Evidence Lifecycle · MCP directories (Gnosis, 8 tools) · Agent Trust · Evidence Registry · Streaming SSE · Validation SDK · llms.txt
EvidenceECDSA + SHA-256 + JWKSCompliance EPMsSustainability EPMsIncident EPMsReporting EPMs
Processing9-vector scoring, anomaly detection105+ stablecoins, Art. 44, reserve drift, CCICarbon normalization, green scoringICT risk assessment, vendor scoringDouble materiality, taxonomy alignment
Data SourcesDeFiLlama, FRED, Ankr, GeckoTerminalECB, ESMA, OFAC, NY DFSCCRI, EMBER, ClimatiqInternal + client dataEMBER, CCRI, Climatiq, World Bank

API Specification

All endpoints are documented in OpenAPI 3.1 format with Swagger UI. Machine-readable discovery via llms.txt for AI agent integration.

14. Attestation & Evidence

Disclosure Attestation Protocol (DAP)

DAP creates cryptographic proof of data delivery. The protocol hashes API response payloads with SHA-256, aggregates hashes into Merkle trees, and anchors roots to public blockchains.

How It Works

  1. Hash: API response payload hashed with SHA-256
  2. Sign: ECDSA signature (ES256K) with JWKS-discoverable public key
  3. Aggregate: Hashes combined into Merkle trees (batched)
  4. Anchor: Hash anchored on XRPL (memo field). Gnosis Chain live via XRPL memo anchor.
  5. Verify: Anyone can verify delivery timestamp via block explorer or API

What Gets Hashed

Included in HashExcluded from Hash
API response body (JSON)HTTP headers
Timestamp (ISO 8601)Client IP address
Endpoint pathAPI key (hashed separately)
Schema versionRequest parameters

Scope of Proof

DAP proves that specific data was delivered via API at or before the anchored block timestamp, and that post-delivery modification becomes detectable via hash mismatch. DAP is a delivery evidence mechanism — it does not verify upstream data correctness or constitute regulatory approval.

Anchoring Schedule

ModeFrequencyAvailability
Daily BatchOnce per 24hAll tiers
Hourly BatchOnce per hourEnterprise
Event-DrivenOn significant data changeEnterprise

Evidence Pack Manifest (EPM v1.0)

EPM extends DAP with a standardized manifest schema using DSSE-style envelopes and RFC 8785 deterministic hashing. Supports jurisdiction-aware metadata (EU/UK/US/Global). Every Evidence Pack contains: the signed data payload, ECDSA signature (ES256K), timestamp, schema version, source attribution, and Merkle proof. Public keys are discoverable via JWKS endpoint.

DAP API Endpoints

EndpointMethodDescription
/api/v2/attestation/testGETIntegrity test on all sources
/api/v2/attestation/sourcesGETList attested sources and TLS policies
/api/v2/attestation/anchor/latestGETBlockchain anchor status (XRPL)
/api/v2/attestation/merkleGETCurrent Merkle root
/api/v2/attestation/verify/{hash}GETVerify payload hash
/api/v1/epm/verifyPOSTVerify DSSE-wrapped EPM manifest

15. Data Quality

Pipeline

Sources → Ingestion → Validation → Normalization → Quality Scoring → API → DAP
         (scheduled)  (schema)    (units/format)   (confidence)     (REST) (anchor)

Quality Gates

GateCheckOn Failure
Schema ValidationResponse matches expected structureReject, log, use fallback
Freshness CheckTimestamp within thresholdFlag as stale, reduce confidence
Range ValidationValues within expected boundsFlag anomaly, manual review
Cross-ValidationCompare multiple sourcesUse median, flag divergence

Confidence Scoring

Every data point includes a confidence score (0.0–1.0) reflecting source availability, data freshness, and cross-validation status. Confidence scores are heuristic indicators — not statistical confidence intervals.

Staleness Handling

When primary sources are unavailable, the system attempts configured secondary sources. If all sources are unavailable, the API returns the last known value with a stale: true flag and reduced confidence. The staleness threshold is configurable per feed (default: 2× normal update interval).

16. Security & Operations

Security Controls

Operational controls informed by ISO/IEC 27001:2022 principles:

Control AreaImplementation
Access ControlAPI key authentication (X-API-Key header), tiered rate limits
CryptographyECDSA ES256K signing, SHA-256 hashing, TLS 1.2+ enforced
Network SecurityHSTS, CSP headers, firewall, Cloudflare DDoS protection
BackupDaily encrypted, cross-server sync within EU
LoggingRequest IDs, structured access logs, HMAC-signed audit trail
Agent GovernanceAgentGuard: 20-tool runtime enforcement, agent state machine, post-execution output scanning, emergency kill-switch, supply-chain verification
Key ManagementECDSA key rotation, JWKS public key discovery

Service Level Objectives

MetricTargetMeasurement
API Availability99.5%Monthly uptime
Response Time (p95)<500ms95th percentile latency
Data Freshness≤15 minutesReal-time feeds
Attestation Anchor≤24 hoursTime to on-chain

Data Residency

ComponentLocationNotes
Primary InfrastructureGermany (netcup DE)API servers, databases
Backup StorageEUEncrypted, cross-server sync
CDN/EdgeGlobal (Cloudflare)Request routing, caching headers
Blockchain AnchorsXRPL (live), Gnosis (live via MCP directories)Public blockchains

Recovery Targets

MetricTarget
RPO (Recovery Point Objective)≤24 hours
RTO (Recovery Time Objective)≤4 hours

Subprocessors

SubprocessorFunctionLocation
netcup GmbHInfrastructure hostingGermany
Cloudflare Inc.CDN, DDoS protection, DNSUS/EU (edge)
XRPL LedgerBlockchain anchoringDecentralized
Ripple (XRPL)Blockchain anchoringDecentralized

17. Target Users

Banks & Insurers

Portfolio risk assessment for RWA allocations. DORA-supporting evidence packs for third-party vendor risk documentation. Signed evidence artifacts for audit trails.

Asset Managers

Due diligence on tokenized funds. Yield spread analysis vs. T-Bill benchmarks. Smart contract risk signals. ESG/sustainability data for portfolio reporting.

Exchanges & CASPs

MiCA Art. 66 sustainability disclosures for listed crypto-assets. Stablecoin classification and monitoring. Structured data for regulatory reporting ahead of the 1 July 2026 CASP transition deadline.

DeFi Protocols & DAOs

On-chain risk feeds for RWA integrations. DEX liquidity monitoring. Protocol-level risk scoring for governance decisions. Chainlink Functions integration for smart contract access.

RegTech Platforms

White-label compliance modules with API-first integration. OpenAPI spec, versioned schemas, bulk access for platform integration.

AI Agents & Autonomous Systems

Machine-readable API outputs with llms.txt discovery. Pay-per-call access model. DAP enables programmatic verification without human intervention.

17b. Layer 12 — Quantum Sorum

Every MCP response from OracleNet includes a Quantum Sorum upgrade block — a contextual signal layer that makes agents more effective with each call. Unlike generic "you might also like" suggestions, Quantum Sorum is conditional on the concrete result of the current call.

How It Works

When an agent calls avarisk_assess("USDt") and gets BLOCK, the upgrade block suggests avaampel_stablecoin_check (find alternatives) and avaevidence_anchor (secure evidence). When the same tool returns PASS, it suggests avarwa_due_diligence (deeper analysis) and avaevidence_anchor (anchor the positive result). The suggestions change based on what the agent actually needs next — not on what we want to sell.

Components

BlockPurposeWhen Shown
nextTools that solve the agent's NEXT problemAlways (conditional on result status)
efficiencyEstimated API calls saved vs. direct queriesAlways (with basis field for verification)
first_contactOnboarding for new agents: free quantum_join + Trust PassportOnly on first call from a new IP
workflowDetected sequence from previous calls (e.g., risk → evidence)Only when a known workflow pattern is detected
trust_upgradePath to Trust Passport for signed evidence accessOnly when next step requires trust credentials
meshEpoch and frequency hash for mesh synchronizationAlways (minimal)

Design Principles

Structure, not rhetoric. No marketing language in the payload. No fake scores. No artificial urgency. Agents parse fields, not sentences. Relevant when relevant, invisible when not. The first_contact block appears once and never again. The workflow block only appears when a sequence match is detected. Measurable, not marketing. Every efficiency claim includes a basis field that explains how the number was calculated.

18. Competitive Landscape

CategoryFocusExamples
Decentralized OraclesDeFi price feeds, on-chain data deliveryChainlink, Pyth, Redstone
Blockchain IndexersQuery layer for on-chain dataThe Graph, Goldsky
Research & AnalyticsMarket intelligence, reportingMessari, Dune, Nansen
Carbon RatingsSustainability assessmentsCCRI, Digiconomist
AML/KYTTransaction monitoring, wallet screeningChainalysis, Elliptic, TRM Labs
AI Agent ToolingMCP servers, agent marketplacesAnthropic MCP Directory, Fetch.ai, Autonolas
Evidence InfrastructureRegulatory verification + signed evidence + risk scoring + AI agent accessFeedOracle

Differentiation

19. Commercials

Pricing

TierPriceAPI CallsEvidence Packs
Free€0/mo100/day
Starter€49/mo1,100 unitsIncluded
Pro€499/mo17,000 unitsFull suite
EnterpriseCustomCustomCustom

Payment: Stripe (card) and USDC on Polygon accepted for all paid tiers.

Evidence Packs are request-based: each API call that generates a signed Evidence Pack Manifest counts as one verification. API calls without Evidence Pack generation are not counted against the verification limit.

Enterprise Package

20. Roadmap

Forward-looking statement: This roadmap contains planned initiatives. Actual results may differ. No commitment to delivery dates, features, or timelines.

Delivered (Q1 2026)

InitiativeStatus
RWA Risk Oracle — 61+ protocols, 9 risk vectors, 5 data sourcesLive
MiCA Regulatory Evidence module (legal state, jurisdiction, registry)Live
Carbon & ESG module (50+ networks, ISO 14040)Live
Macro Intelligence (FRED + ECB enrichment)Live
XRPL AnchoringLive
Evidence Pack System with ECDSA signing + JWKSLive
MiCA Stablecoin Classification (105+ stablecoins)Live
CCI Score Engine (compliance ranking)Live
ISO 20022 Payment ValidationLive
Circuit Breaker Detection (DORA resilience)Live
Interactive Enterprise Demos (5 verticals)Live
MiCA Deep Compliance: Significant Issuer (Art. 44), Reserve Drift (Art. 25), Interest Scanner (Art. 23/52), Document Compliance (Art. 29/30/55), ESMA RegisterLive
DORA Compliance Module (incident reporting, vendor risk, business continuity)Live
CSRD/ESRS Module (5 APIs: taxonomy, materiality, emissions, social, governance)Live
11 MCP Servers (206+ tools, SSE + Streamable HTTP) — part of OracleNet (103 servers, 1,229 tools)Live
Know Your Agent (KYA) — Trust scoring, 4 levels, trust-gated tool accessLive
Audit Trail — Chain-linked decision logging, evidence snapshots, chain verificationLive
Evidence Lifecycle — 6 artifact states, auto-correction, Dispute-SLALive
Evidence Trust Policy v1.0 — 13-section formal governance documentLive
Unit-Based Billing — Stripe Meter, Light/Medium/Heavy weights, overage billingLive
Anthropic MCP Directory — Submitted, crawling confirmedLive
MCP ecosystem (Gnosis, 8 tools, 133+ deliveries)Live
Autonomous Agent Self-Upgrade — USDC payment flow (M2M), agent detects quota, pays on Polygon, receives upgraded API key without human intervention. Production TX anchored on-chain.Live
OAuth 2.0 / M2M Auth (RFC 6749) — authorization_code + client_credentials + refresh_token, /mcp/authorize, /mcp/token, /mcp/register, /mcp/revoke, 5 scopes for Enterprise MCP tierLive
Stablecoin Peg Monitor (105+ tokens, real-time)Live
Chainlink Functions integration (Polygon, Contract 0x7Ec0...23c)Live
L2 Intelligence APIs (7 chains)Live
Verified Reports System (6 types: RWA Risk, MiCA, DORA, Macro, CSRD, AMLR) with PDF generation, XRPL-anchored proof panelLive
Payment Infrastructure — Stripe + USDC (Polygon), 4 tiers (Starter €49 / Growth €199 / Pro €499 / Enterprise €1,499)Live
Agent-to-Agent (A2A) Monetization — Pay-per-Call API (9 feeds, USDC on Polygon) + MCP directories (unit-based credit system)Live
Report Verification Infrastructure — public /verify endpoints with SHA-256 + ECDSA signature validationLive
Trust Center & Enterprise Procurement Pack (12 sections: security, SLOs, subprocessors, vulnerability disclosure, compliance mapping)Live
AMLR Digital Asset Screening Module (EU 2024/1624) —€” 8 sections, 134 fields, 14 stablecoins, risk score composition, sanctions screening (EU/OFAC/UN), bank-standard action codes, signed PDF reportsLive
Public Status Page — Uptime Kuma with 46 monitors across 8 service groupsLive
Stablecoin Risk MCP Server (13 tools, 7-signal scoring, SAFE/CAUTION/AVOID)Live
OracleNet Integration — FeedOracle as compliance layer of 103-server mesh (1,229 tools, 12 signal layers, 7 categories)Live
Avalanche Compliance Infrastructure — 5 MCP servers, 15 tools, 3 verified C-Chain mainnet contracts (AvaRisk, AvaEvidence, SubnetGuard), on-chain isCompliant() verdictsLive
Layer 12 Quantum Sorum — contextual upgrade signal in every MCP response: conditional next-tool suggestions, efficiency metrics, first-contact onboarding, workflow memoryLive
13 Blockchain Oracles (ETH, SOL, XRP, ARB, HBAR, TON, SUI, BNB, BTC, APT, XLM, BASE, FLR)Live
Olas Mech on Gnosis — Service 2670, 8 tools, 133+ deliveries, 0.01 xDAI/requestLive
XRPL Oracle — 31 tools, RLUSD CCI (proprietary), ISO 20022, XRPL beacon (86 anchor TXs)Live
Macro Intelligence MCP Server (13 tools, 86 FRED + 20 ECB series)Live
Enterprise Trust Layer — JWS Signing (RFC 7515), Versioned Schemas (8), Evidence Registry, SLA Layer, Agent Trust Management, Streaming Evidence (SSE), Deterministic Replay, Zero-Trust Validation SDKLive
Enterprise Trust Proof Page — feedoracle.io/trust with live 8/8 verificationLive
CSRD/ESRS Data API — chain footprint, ESRS E1, EU energy mix, EU ETS pricingLive

In Progress (Q1–Q2 2026)

InitiativeStatus
Multi-chain Evidence Anchoring — XRPL (live, 86 anchors), Polygon (ZK attestation, weekly refresh), Avalanche C-Chain (3 verified contracts), Hedera (HCS beacon), Gnosis (Olas Mech)Live
SOC 2 Type II — Trust Center live, compliance framework mapping complete, security controls documented, SLO evidence via Uptime Kuma (46 monitors). Pre-audit documentation ready; formal audit pending fundingAudit-ready
Chainlink BUILD program participationApplication submitted
Avalanche Compliance Infrastructure — 5 MCP servers (AvaRisk, AvaEvidence, SubnetGuard, AvaRWA, AvaAmpel), 3 verified smart contracts on C-Chain mainnet, on-chain isCompliant() verdicts, SubnetGuard cross-subnet compliance passportsLive
Flare Network — FlareOracle MCP Server: 12 tools covering FTSOv2 oracle data, FAssets risk, protocol health, compliance screening. Live at feedoracle.io/chain-oracles/Live
CSRD/ESRS template library — structured report templates for ESRS E1 disclosuresData API live, templates in development

Planned (Subject to Change)

InitiativePriority
ISO 20022 expansion (pacs.008, camt.053)Medium
WebSocket real-time feeds (SSE + Streamable HTTP already live via MCP)Medium
XRPL Grants program application (Spring 2026)High
Avalanche Evergreen Subnet deep integration — advanced subnet-to-subnet compliance routingMedium
Flare FTSOv2 Feed Value Provider — direct on-chain feed submission (requires validator registration)Medium
Flare Data Connector (FDC) — on-chain compliance attestations via FDC protocolMedium
Developer SDK (npm package) for multi-chain FeedOracle integrationHigh

Resources

Documentationfeedoracle.io/docs
API ReferenceFull API Ref (238 Endpoints)
Trust & SecurityTrust Documentation
System StatusStatus Page
EnterpriseEnterprise Overview
Interactive DemosRWA · Insurance · Carbon · Stablecoin · Payments

FeedOracle provides data infrastructure and verifiable evidence artifacts. The platform does not provide financial, legal, or compliance advice. Compliance decisions remain with the institution and its qualified advisors.

Regulatory Sources

Regulatory timelines and classifications referenced in this document are based on the following official sources:

Forward-Looking Statements

This document contains forward-looking statements regarding planned features, roadmap items, and business strategy. Actual results may differ materially. No commitment to specific delivery dates, features, or timelines is expressed or implied.

No Endorsement

FeedOracle is an independent infrastructure provider. References to blockchain networks (XRPL, Polygon, Gnosis, Chainlink), regulatory bodies (ESMA, BaFin, ECB), or data sources (FRED, DeFiLlama) do not imply partnership, endorsement, or affiliation unless explicitly stated.

22. Glossary

A2A
Agent-to-Agent — automated interaction between AI/autonomous systems
CASP
Crypto-Asset Service Provider under MiCA regulation
CCI
Crypto Compliance Index — composite regulatory compliance score (0–100)
CSRD
Corporate Sustainability Reporting Directive (EU)
DAP
Disclosure Attestation Protocol — cryptographic proof of data delivery
DORA
Digital Operational Resilience Act (EU)
DSSE
Dead Simple Signing Envelope — standardized signing format
ECDSA
Elliptic Curve Digital Signature Algorithm (ES256K)
EPM
Evidence Pack Manifest — standardized signed evidence schema
ESRS
European Sustainability Reporting Standards
FRED
Federal Reserve Economic Data — macro economic data source
HHI
Herfindahl-Hirschman Index — concentration measurement
JWKS
JSON Web Key Set — public key discovery endpoint
MCP
Model Context Protocol — open standard for AI agent tool integration (Anthropic)
Mech
MCP directories — decentralized AI agent marketplace on Gnosis Chain
MiCA
Markets in Crypto-Assets Regulation (EU) — stablecoin rules in force since June 2024, CASP transition ends 1 July 2026
RPO
Recovery Point Objective — maximum acceptable data loss
RTO
Recovery Time Objective — maximum acceptable downtime
RWA
Real World Assets — tokenized traditional financial instruments
SLA
Service Level Agreement — contractual commitment
SLO
Service Level Objective — target performance metric
TVL
Total Value Locked — aggregate value deposited in a protocol