Atomadic starter MCP (50 tools)

atomadictech/atomadic-mcp

Formally verified safety infrastructure for autonomous agents. Exposes 75+ endpoints covering cryptographic randomness (VeriRand), hallucination detection with certified probability bounds, session credential rotation that addresses CVE-2025-6514 (RatchetGate), UCAN delegation chains with proven depth limits, and A2A economy primitives like escrow and reputation. Pay per call via x402 USDC micropayments starting at $0.008, no signup required. Ships with 872 verified theorems in the proof chain. Reach for this when you're building agent-to-agent pipelines that handle money, sign contracts, or make decisions without human oversight and need mathematical guarantees instead of benchmark scores.

tools: 50
transport: HTTP
last updated: Oct 27, 2026

View on GitHub

Install

From registry metadata

$ npx add-mcp https://mcp.atomadic.tech/mcp/public --name aaaa-nexus

Run in your terminal. add-mcp asks which agents to add it to (Claude Code, Codex, Cursor and more). Add -g to install it for every project.

Review the command, arguments and environment values before installing. MCP servers run with your local permissions.

Other ways

Claude Code

$ claude mcp add --transport http aaaa-nexus https://mcp.atomadic.tech/mcp/public

Run in your terminal. Add --scope user to make it available in every project.

Tools

Verified live · 50 tools

Verified live against the running server on Nov 6, 2026.

murmurationSTART HERE (free, no key). Initialize your Atomadic control surface in one call: returns a verbose, step-by-step setup (SDK install, MCP connect, your unlocked tools), a RUNNABLE polyglot demo that proves the superpowers, recon links, and a one-command bootstrap. Idempotent —...1 params

START HERE (free, no key). Initialize your Atomadic control surface in one call: returns a verbose, step-by-step setup (SDK install, MCP connect, your unlocked tools), a RUNNABLE polyglot demo that proves the superpowers, recon links, and a one-command bootstrap. Idempotent —...

Parameters* required

agent_idstring

Optional stable id to track your onboarding state across calls.

agents_contradiction[agents] Detect contradictions between two free-text agent outputs (natural-language claim drift). For structured key/value claims that must agree, use detect_contradiction instead.1 params

[agents] Detect contradictions between two free-text agent outputs (natural-language claim drift). For structured key/value claims that must agree, use detect_contradiction instead.

Parameters* required

statements*array

agents_semantic_diff[agents] Semantic diff between two free-text agent outputs (meaning-level change, not text-level). Use forge_verify for build artifacts and drift_check for certified operational drift.3 params

[agents] Semantic diff between two free-text agent outputs (meaning-level change, not text-level). Use forge_verify for build artifacts and drift_check for certified operational drift.

Parameters* required

contextstring

output_a*string

output_b*string

agents_register[agents] Register an agent2 params

[agents] Register an agent

Parameters* required

agent_id*string

capabilities*array

agents_reputation[agents] Agent reputation and trust score1 params

[agents] Agent reputation and trust score

Parameters* required

agent_id*string

agents_plan[agents] Autonomous agent goal planning2 params

[agents] Autonomous agent goal planning

Parameters* required

constraintsobject

Optional constraints for the plan (e.g., time limits, resource caps)

goal*string

reputation_record[agents] Record agent interaction4 params

[agents] Record agent interaction

Parameters* required

agent_id*string

counterparty_idstring

interaction_type*string

outcome*string

one of positive · negative · neutral

sla_register[agents] Register an SLA3 params

[agents] Register an SLA

Parameters* required

consumer_id*string

provider_id*string

terms*object

sla_report[agents] Report SLA violation3 params

[agents] Report SLA violation

Parameters* required

descriptionstring

sla_id*string

violation_type*string

swarm_inbox[agents] Poll swarm inbox

[agents] Poll swarm inbox

No parameters. Call it with no arguments.

swarm_relay[agents] Relay message through swarm3 params

[agents] Relay message through swarm

Parameters* required

message*object

target_agent_id*string

ttlinteger

Hop limit for relay

compliance_check[security] Delegation-depth + KL gate (GOV-100). Reads chain_depth and kl_divergence. It does NOT read your data description and does NOT evaluate GDPR/CCPA - compliant:true only means the delegation chain is shallow and KL is under epsilon.3 params

[security] Delegation-depth + KL gate (GOV-100). Reads chain_depth and kl_divergence. It does NOT read your data description and does NOT evaluate GDPR/CCPA - compliant:true only means the delegation chain is shallow and KL is under epsilon.

Parameters* required

chain_depthinteger

Delegation chain depth. Higher = deeper hand-off. Past the gate threshold the result escalates to quarantine_escalation.

descriptionstring

Free text. Accepted for readability and IGNORED by the gate. It never changes the verdict.

kl_divergencenumber

eps_KL divergence of the current policy from baseline. Above epsilon fails the gate.

compliance_eu_ai_act[security] EU AI Act compliance certificate3 params

[security] EU AI Act compliance certificate

Parameters* required

intended_purpose*string

risk_category*string

one of minimal · limited · high · unacceptable

system_id*string

compliance_explain[security] Explainability certificate (XPL-100-ExplainBound)3 params

[security] Explainability certificate (XPL-100-ExplainBound)

Parameters* required

decision_id*string

model_id*string

subject_context*object

compliance_fairness[security] Fairness proof (FNS-100-FairnessBound)4 params

[security] Fairness proof (FNS-100-FairnessBound)

Parameters* required

dataset_id*string

metric*string

model_id*string

protected_attrs*array

compliance_lineage[security] Data lineage (LIN-100-ChainIntegrity)2 params

[security] Data lineage (LIN-100-ChainIntegrity)

Parameters* required

dataset_stages*array

model_id*string

compliance_transparency[security] Transparency report3 params

[security] Transparency report

Parameters* required

include*array

period*string

system_id*string

compliance_incident[security] Report compliance incident4 params

[security] Report compliance incident

Parameters* required

affected_users*integer

description*string

severity*string

one of low · medium · high · critical

system_id*string

compliance_incidents[security] Incident registry1 params

[security] Incident registry

Parameters* required

system_id*string

compliance_oversight[security] Log oversight event4 params

[security] Log oversight event

Parameters* required

decision*string

notes*string

reviewer_id*string

system_id*string

security_zero_day[security] Zero-day threat assessment2 params

[security] Zero-day threat assessment

Parameters* required

contextobject

target*string

threat_score[security] Multi-vector threat score2 params

[security] Multi-vector threat score

Parameters* required

agent_id*string

contextstring

defi_optimize[defi] LP parameter optimization (DFP-100-OptimalTick)3 params

[defi] LP parameter optimization (DFP-100-OptimalTick)

Parameters* required

position*object

protocol*string

risk_tolerance*number

defi_risk_score[defi] DeFi risk scoring (DFI-101-RiskScoreBound)3 params

[defi] DeFi risk scoring (DFI-101-RiskScoreBound)

Parameters* required

market_conditions*object

position*object

protocol*string

defi_yield_optimize[defi] Yield Oracle (YLD-100-YieldConvergence)2 params

[defi] Yield Oracle (YLD-100-YieldConvergence)

Parameters* required

principal_usd*number

risk_tolerance*number

escrow_create[defi] Create agent-to-agent escrow5 params

[defi] Create agent-to-agent escrow

Parameters* required

amount*number

conditionsobject

currency*string

payee_id*string

payer_id*string

escrow_release[defi] Release escrowed funds1 params

[defi] Release escrowed funds

Parameters* required

escrow_id*string

escrow_dispute[defi] Open escrow dispute3 params

[defi] Open escrow dispute

Parameters* required

escrow_id*string

evidenceobject

reason*string

forge_badge[fuse] Model quality badge1 params

[fuse] Model quality badge

Parameters* required

model*string

Model id to fetch the quality badge for.

forge_leaderboard[fuse] Model leaderboard

[fuse] Model leaderboard

No parameters. Call it with no arguments.

forge_verify[fuse] Verify a model submission2 params

[fuse] Verify a model submission

Parameters* required

model_id*string

verification_typestring

forge_delta_submit[fuse] Submit a model delta3 params

[fuse] Submit a model delta

Parameters* required

delta*object

descriptionstring

model_id*string

identity_verify[identity] Topological identity verification1 params

[identity] Topological identity verification

Parameters* required

agent_id*string

identity_sybil_check[identity] Sybil-resistance check

[identity] Sybil-resistance check

No parameters. Call it with no arguments.

ratchet_register[identity] RatchetGate â€” register session1 params

[identity] RatchetGate â€” register session

Parameters* required

session_id*string

ratchet_advance[identity] RatchetGate â€” advance session1 params

[identity] RatchetGate â€” advance session

Parameters* required

session_id*string

inference[inference] Chat inference with hallucination guard1 params

[inference] Chat inference with hallucination guard

Parameters* required

messages*array

inference_stream[inference] Streaming inference with chain-of-thought1 params

[inference] Streaming inference with chain-of-thought

Parameters* required

messages*array

rag_augment[inference] Trusted RAG augmentation

[inference] Trusted RAG augmentation

No parameters. Call it with no arguments.

text_summarize[inference] Text summarization1 params

[inference] Text summarization

Parameters* required

text*string

rng_quantum[inference] Quantum random number with proof

[inference] Quantum random number with proof

No parameters. Call it with no arguments.

rng_verify[inference] Verify current epoch random value

[inference] Verify current epoch random value

No parameters. Call it with no arguments.

trust_score[trust] Bounded trust score1 params

[trust] Bounded trust score

Parameters* required

agent_id*string

drift_check[trust] DriftGuard check (DRG-100-DriftBound)3 params

[trust] DriftGuard check (DRG-100-DriftBound)

Parameters* required

current_window*object

model_id*string

reference_window*object

drift_certificate[trust] Drift certificate1 params

[trust] Drift certificate

Parameters* required

id*string

oracle_hallucination[trust] Hallucination oracle2 params

[trust] Hallucination oracle

Parameters* required

claim*string

contextstring

audit_work_claim[Ledger product, entitlement: ledger] Recompute-audit an agent work claim: re-derive a deterministic SHA-256 digest from the claimed inputs and check it against the claimed digest. Returns GREEN (matches), RED (tampered), or UNVERIFIED (no claim supplied). Replay, not opinion....2 params

[Ledger product, entitlement: ledger] Recompute-audit an agent work claim: re-derive a deterministic SHA-256 digest from the claimed inputs and check it against the claimed digest. Returns GREEN (matches), RED (tampered), or UNVERIFIED (no claim supplied). Replay, not opinion....

Parameters* required

claimed_digeststring

The agent's claimed SHA-256 digest to verify against. Omit for UNVERIFIED + the recomputed digest.

inputs*object

The work/claim payload to recompute a digest over.

verify_settlement[Ledger product, entitlement: ledger] Proof-gated settlement: recompute the work digest and authorize a payout only if the claimed digest verifies GREEN. No proof, no payout. 3 free calls/day, then entitlement 'ledger'.2 params

[Ledger product, entitlement: ledger] Proof-gated settlement: recompute the work digest and authorize a payout only if the claimed digest verifies GREEN. No proof, no payout. 3 free calls/day, then entitlement 'ledger'.

Parameters* required

claimed_digeststring

Claimed digest the payout is conditioned on.

inputs*object

The work payload backing the settlement.

detect_contradiction[Ledger product, entitlement: ledger] Detect contradictions across structured claims: the same key asserted with conflicting values flips RED, otherwise GREEN. Deterministic and recomputable. Use agents_contradiction instead when comparing free-text agent prose. 3 free calls/d...1 params

[Ledger product, entitlement: ledger] Detect contradictions across structured claims: the same key asserted with conflicting values flips RED, otherwise GREEN. Deterministic and recomputable. Use agents_contradiction instead when comparing free-text agent prose. 3 free calls/d...

Parameters* required

claims*array

List of claim objects, each with a 'key' (or 'subject') and a 'value'.

render_polyglot_all_pure[Fuse product · entitlement: fuse · FREE_R2_DEMO] Polyglot emit to all 7 supported targets in one call. Returns {ok, ir, results, languages}. Canonical demo examples baked to R2; novel inputs via SDK.2 params

[Fuse product · entitlement: fuse · FREE_R2_DEMO] Polyglot emit to all 7 supported targets in one call. Returns {ok, ir, results, languages}. Canonical demo examples baked to R2; novel inputs via SDK.

Parameters* required

atom_name*string

source_text*string

Readme

aaaa-nexus — public entry point

This repo is the public front page for the AAAA-Nexus hosted MCP server: formally verified AI safety and compliance APIs, reachable over HTTP with no signup.

The server is live. The source is not public. That is the whole point of this repo: if you landed here from an MCP directory and every other Atomadic link was a dead end, this is the one that resolves.

Live endpoint

https://aaaa-nexus.atomadictech.workers.dev/mcp

Streamable HTTP. The server negotiates down to whatever revision the client asks for (a 2025-06-18 probe is answered with 2024-11-05). Verified live:

curl -s -X POST https://aaaa-nexus.atomadictech.workers.dev/mcp \
  -H 'content-type: application/json' \
  -H 'accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"probe","version":"1"}}}'

Then tools/list and tools/call. 44 tools are exposed. A sample:

rng_quantum, threat_score, hallucination_oracle, identity_verify, ratchet_register, authorize_action, spending_authorize, lineage_record.

View the full README on GitHub

Details

Repository: github.com/atomadictech/atomadic-mcp
Website: atomadic.tech
Transport: HTTP
Registry: MCP registry, active
Tools verified: Nov 6, 2026