Back to AAAA-Nexus
Card snapshot
atomadic.tech
·
2026-05-15 00:30:11 UTC
·
3a1c3f3932e9c03fc9a2e492f7b3aeb418624bfca0ae2d56c5f1d321b14d4454
This is a frozen copy of the agent's agent-card.json as we observed it at the timestamp above. We capture a new snapshot every time the card's content hash changes. Useful for: forensic drift analysis, verifying downstream callers see the right version, reproducing routing decisions made historically.
{
"authentication": {
"credentials": null,
"schemes": [
"Bearer",
"APIKey",
"x402"
],
"trialAccess": "3 calls/endpoint/day with no credentials \u2014 real production responses"
},
"capabilities": {
"freeTrial": true,
"freeTrialCallsPerEndpointPerDay": 3,
"pushNotifications": false,
"stateTransitionHistory": false,
"streaming": false,
"x402Payment": true
},
"defaultInputModes": [
"application/json"
],
"defaultOutputModes": [
"application/json"
],
"description": "Agent Control Plane - 146+ endpoints for AI agent security, trust, reputation, escrow, SLA enforcement, formal verification, compliance, discovery, and x402 USDC micropayments on Base L2. Deterministic pre-action authorization, Lean 4 verified safety constants, and tamper-proof decision audit trails.",
"documentationUrl": "https://atomadic.tech/quickstart",
"name": "AAAA-Nexus",
"preferredTransport": "JSONRPC",
"protocolVersion": "1.0.0",
"provider": {
"organization": "Atomadic Tech",
"url": "https://atomadic.tech"
},
"skills": [
{
"description": "Quantum-seeded random bytes with HMAC-SHA256 proof of entropy",
"examples": [
"Generate 32 cryptographically secure random bytes",
"Verify entropy proof for a previous RNG output"
],
"id": "quantum-rng",
"inputModes": [
"application/json"
],
"name": "Quantum RNG",
"outputModes": [
"application/json"
],
"tags": [
"randomness",
"entropy",
"cryptography",
"free"
]
},
{
"description": "Score a payload for security threats using triality consensus",
"examples": [
"Score this JSON payload for injection threats"
],
"id": "threat-score",
"inputModes": [
"application/json"
],
"name": "Threat Scoring",
"outputModes": [
"application/json"
],
"tags": [
"security",
"threat",
"compliance"
]
},
{
"description": "47-epoch safe-prime session ratchet with formal re-key schedule (MCP CVE-2025-6514 mitigation). Register, advance, probe, and query session status.",
"examples": [
"Register a new 47-epoch session ratchet",
"Advance session to next epoch",
"Probe ratchet health"
],
"id": "ratchetgate",
"inputModes": [
"application/json"
],
"name": "RatchetGate Session Security",
"outputModes": [
"application/json"
],
"tags": [
"security",
"session",
"cryptography",
"mcp"
]
},
{
"description": "LLM inference via Cloudflare Workers AI \u2014 Llama 3.1 8B with streaming support",
"examples": [
"Run inference on Llama 3.1 8B",
"Stream a completion response"
],
"id": "ai-inference",
"inputModes": [
"application/json"
],
"name": "AI Inference",
"outputModes": [
"application/json"
],
"tags": [
"inference",
"llm",
"ai"
]
},
{
"description": "Multi-agent escrow for high-value A2A transactions \u2014 create, release, dispute, and arbitrate escrow contracts",
"examples": [
"Create an escrow for a 50 USDC agent-to-agent task",
"Release escrow after task verification",
"Dispute an escrow outcome"
],
"id": "escrow",
"inputModes": [
"application/json"
],
"name": "Agent Escrow",
"outputModes": [
"application/json"
],
"tags": [
"escrow",
"payments",
"trust",
"arbitration"
]
},
{
"description": "Decentralized reputation registry \u2014 record interactions, query trust scores, view history, and dispute ratings. Cross-references x402 payment history for Sybil resistance.",
"examples": [
"Record a successful interaction with agent X",
"Get the trust score for agent Y",
"Dispute a negative reputation entry"
],
"id": "reputation",
"inputModes": [
"application/json"
],
"name": "Agent Reputation Oracle",
"outputModes": [
"application/json"
],
"tags": [
"reputation",
"trust",
"sybil-resistance",
"a2a"
]
},
{
"description": "Register, monitor, and enforce service-level agreements between agents \u2014 automatic breach detection and reporting",
"examples": [
"Register an SLA with 99.5% uptime requirement",
"Report SLA metrics",
"Query breach status"
],
"id": "sla-enforcement",
"inputModes": [
"application/json"
],
"name": "SLA Enforcement",
"outputModes": [
"application/json"
],
"tags": [
"sla",
"monitoring",
"compliance",
"a2a"
]
},
{
"description": "Search the agent registry, get capability-matched recommendations, and browse the global agent directory",
"examples": [
"Find agents that provide image generation",
"Recommend agents for a multi-step research task"
],
"id": "discovery",
"inputModes": [
"application/json"
],
"name": "Agent Discovery & Matchmaking",
"outputModes": [
"application/json"
],
"tags": [
"discovery",
"registry",
"matchmaking",
"a2a"
]
},
{
"description": "EU AI Act compliance checks, ethics screening, prompt injection scanning, and AI Bill of Materials drift detection",
"examples": [
"Check EU AI Act compliance for an agent card",
"Scan a prompt for injection attempts",
"Detect AIBOM drift"
],
"id": "compliance",
"inputModes": [
"application/json"
],
"name": "Compliance & Ethics",
"outputModes": [
"application/json"
],
"tags": [
"compliance",
"eu-ai-act",
"ethics",
"aibom",
"safety"
]
},
{
"description": "Topological identity verification, Sybil-check, zero-trust attestation, and UCAN-style delegation validation",
"examples": [
"Verify an agent's topological identity",
"Validate a delegation chain",
"Run a Sybil resistance check"
],
"id": "identity",
"inputModes": [
"application/json"
],
"name": "Identity & Delegation",
"outputModes": [
"application/json"
],
"tags": [
"identity",
"verification",
"delegation",
"pqc",
"zero-trust"
]
},
{
"description": "Tamper-proof audit logging and verification \u2014 capture full decision cycles with RatchetGate-secured integrity proofs for regulatory compliance (GDPR/CCPA)",
"examples": [
"Log an agent decision with reasoning trace",
"Verify audit log integrity"
],
"id": "audit",
"inputModes": [
"application/json"
],
"name": "Decision Audit Vault",
"outputModes": [
"application/json"
],
"tags": [
"audit",
"logging",
"compliance",
"observability"
]
},
{
"description": "LLM hallucination upper-bound constant \u2014 deterministic threshold for output reliability gating",
"examples": [
"Get the hallucination threshold for gating agent outputs"
],
"id": "hallucination-oracle",
"inputModes": [
"application/json"
],
"name": "Hallucination Oracle",
"outputModes": [
"application/json"
],
"tags": [
"hallucination",
"safety",
"oracle",
"llm"
]
},
{
"description": "proprietary geometric model and data compression with decompression \u2014 lossless via E8 lattice geometry",
"examples": [
"Compress a model checkpoint",
"Decompress a HELIX-encoded payload"
],
"id": "compression",
"inputModes": [
"application/json"
],
"name": "HELIX Compression",
"outputModes": [
"application/json"
],
"tags": [
"compression",
"optimization",
"geometric-compression"
]
},
{
"description": "Semantic diff, intent classification, contradiction detection, token budget optimization, capability matching, and multi-step plan generation",
"examples": [
"Classify the intent of an agent message",
"Detect contradictions in an agent plan",
"Generate a multi-step execution plan"
],
"id": "agent-reasoning",
"inputModes": [
"application/json"
],
"name": "Agent Reasoning Tools",
"outputModes": [
"application/json"
],
"tags": [
"reasoning",
"planning",
"nlp",
"a2a"
]
},
{
"description": "Multi-agent swarm message relay and inbox for coordinated agent-to-agent communication",
"examples": [
"Relay a task message to a swarm of agents",
"Check the swarm inbox for pending messages"
],
"id": "swarm-relay",
"inputModes": [
"application/json"
],
"name": "Swarm Relay",
"outputModes": [
"application/json"
],
"tags": [
"swarm",
"messaging",
"coordination",
"a2a"
]
},
{
"description": "Scan agent code or payloads for zero-day vulnerability patterns",
"examples": [
"Scan an agent payload for known zero-day patterns"
],
"id": "zero-day-scanner",
"inputModes": [
"application/json"
],
"name": "Zero-Day Scanner",
"outputModes": [
"application/json"
],
"tags": [
"security",
"zero-day",
"vulnerability"
]
},
{
"description": "Deterministic pre-action authorization for tool calls \u2014 evaluates agent actions against formally verified policy bounds before execution. Returns cryptographic authorization tokens.",
"examples": [
"Authorize a fund transfer action for agent-324",
"Check if agent can invoke a high-risk tool at delegation depth 3"
],
"id": "pre-action-auth",
"inputModes": [
"application/json"
],
"name": "Pre-Action Authorization Gateway",
"outputModes": [
"application/json"
],
"tags": [
"authorization",
"policy",
"governance",
"oap",
"control-plane"
]
},
{
"description": "Agent chain spending authorization with trust-decay bounds \u2014 validates whether an agent or multi-hop delegation chain is authorized to spend a given amount. Prevents runaway spending.",
"examples": [
"Authorize a 50,000 micro-USDC spend at chain depth 2",
"Validate a 5-hop agent chain spending budget"
],
"id": "spending-governance",
"inputModes": [
"application/json"
],
"name": "Spending Governance Oracle",
"outputModes": [
"application/json"
],
"tags": [
"spending",
"governance",
"payments",
"x402",
"control-plane"
]
},
{
"description": "Cross-framework structured decision trace capture \u2014 records intent, reasoning, constraints, alternatives, and outcomes in a tamper-proof hash chain. GDPR/CCPA audit-ready.",
"examples": [
"Record a decision with full reasoning trace",
"Trace a decision chain back to its root"
],
"id": "decision-lineage",
"inputModes": [
"application/json"
],
"name": "Decision Lineage Vault",
"outputModes": [
"application/json"
],
"tags": [
"observability",
"lineage",
"audit",
"compliance",
"control-plane"
]
},
{
"description": "Verify agent behavioral specs against formally proven bounds \u2014 attestation that an agent's stated policies (delegation depth, hallucination rate, spending limits) are within provably safe parameters.",
"examples": [
"Verify an agent's behavioral contract claims",
"Retrieve a Nexus-Certified attestation"
],
"id": "behavioral-contract",
"inputModes": [
"application/json"
],
"name": "Behavioral Contract Verifier",
"outputModes": [
"application/json"
],
"tags": [
"verification",
"contracts",
"formal-methods",
"attestation",
"control-plane"
]
},
{
"description": "Mint platform-neutral agent identity tokens portable across AWS Bedrock, Google ADK, LangChain, CrewAI, AutoGen, and OpenAI Agents SDK. Solves cross-platform agent identity fragmentation.",
"examples": [
"Mint a federated identity for a Google ADK agent",
"Verify a federation token from another platform",
"Check capability portability from AWS to LangChain"
],
"id": "identity-federation",
"inputModes": [
"application/json"
],
"name": "Agent Identity Federation",
"outputModes": [
"application/json"
],
"tags": [
"identity",
"federation",
"portability",
"cross-platform",
"control-plane"
]
},
{
"description": "Coordinate N agents voting on an output with majority/unanimous/weighted/supermajority quorum modes. Issues a signed certificate when consensus is reached. Fills the missing A2A v1.0 consensus primitive.",
"examples": [
"Open a majority-quorum session for 3 agents",
"Cast a vote with confidence 0.92",
"Retrieve the certified consensus result"
],
"id": "consensus-broker",
"inputModes": [
"application/json"
],
"name": "Multi-Agent Consensus Broker",
"outputModes": [
"application/json"
],
"tags": [
"consensus",
"voting",
"coordination",
"multi-agent",
"certification"
]
},
{
"description": "Define a token/call budget tree for an orchestrator and sub-agents. Enforces parent-pool limits with per-child soft caps and idempotency-key retry deduplication to prevent storm-driven budget exhaustion.",
"examples": [
"Create a 500k token tree for root + 3 sub-agents",
"Draw 4000 tokens with idempotency key",
"Check tree status and soft-cap alerts"
],
"id": "quota-ledger",
"inputModes": [
"application/json"
],
"name": "Hierarchical Quota Ledger",
"outputModes": [
"application/json"
],
"tags": [
"quota",
"rate-limiting",
"budget",
"orchestration",
"governance"
]
},
{
"description": "Issue a signed certificate proving an agent output passed a quality rubric before delivery. Built-in rubrics: contract-completeness-v2, safety-review-v1, hallucination-bound-v1. Verifiable offline.",
"examples": [
"Certify a contract draft against completeness rubric",
"Verify an output certificate offline"
],
"id": "output-certifier",
"inputModes": [
"application/json"
],
"name": "Output Certification Notary",
"outputModes": [
"application/json"
],
"tags": [
"certification",
"quality",
"notary",
"attestation",
"output"
]
},
{
"description": "Saga-as-a-Service for multi-step agent workflows \u2014 register compensating actions per step, checkpoint completions, and trigger reverse rollback on failure. Fills the standardized compensation gap.",
"examples": [
"Register a 5-step workflow with compensating actions",
"Checkpoint a completed step",
"Trigger rollback from a failed step"
],
"id": "rollback-coordinator",
"inputModes": [
"application/json"
],
"name": "Agent Rollback Coordinator",
"outputModes": [
"application/json"
],
"tags": [
"rollback",
"saga",
"compensation",
"transactions",
"fault-tolerance"
]
},
{
"description": "Cryptographic namespace boundary preventing cross-tenant memory bleed in warm-pooled agents. HMAC-derived per-tenant keys rotate each epoch. Audit trail for memory access patterns. GDPR/HIPAA audit-ready.",
"examples": [
"Create a strict isolation fence for a tenant session",
"Audit memory access counts",
"Close and purge a fence on session end"
],
"id": "memory-fence",
"inputModes": [
"application/json"
],
"name": "Cross-Tenant Memory Fence",
"outputModes": [
"application/json"
],
"tags": [
"memory",
"isolation",
"multi-tenant",
"privacy",
"security"
]
}
],
"trialPolicy": {
"authRequired": false,
"freeCalls": 3,
"note": "Every paid endpoint allows 3 free trial calls per day with no API key or payment. Responses are identical to production \u2014 this is a proof-of-quality guarantee, not a sandbox. Trial calls include a 2-second rate-limit delay to prevent abuse; paid calls (API key or x402) have zero added latency. After 3 calls, pay per call via x402 USDC or get an API key at https://atomadic.tech/pay",
"period": "daily",
"scope": "per_endpoint_per_ip"
},
"url": "https://atomadic.tech/a2a",
"version": "0.5.1"
}