Atomadic starter MCP (50 tools)

io.github.atomadictech由 GitHub 帳號 atomadictech 發布v0.5.2更新於 Oct 29, 2026

50 starter MCP tools for AI compliance and trust: identity, escrow, SLA, EU AI Act.

發布方自己的服務:Atomadic MCP

已驗證Streamable HTTP可網頁執行金融安全與監控AI 與機器學習

概覽

AI 產生的概覽

Atomadic 託管 MCP 端點,提供 50 個入門工具,涵蓋 AI 合規、代理信任、身分、託管與推論,需付費 API 金鑰解鎖。

可以做什麼
  • 執行免費導引呼叫,回傳安裝步驟、可執行示範以及你的方案已解鎖的工具
  • 檢查合規與治理:GDPR/CCPA 閘門、EU AI Act、可解釋性、公平性、資料血緣、透明度與事件登錄
  • 評估安全與威脅:零日威脅評估與多向量威脅評分
  • 管理代理:註冊、信譽、目標規劃、矛盾偵測與語意差異、SLA 報告、群體收件匣與中繼
  • 處理 DeFi 與託管:LP 與收益最佳化、風險評分,以及建立、釋放或爭議代理間託管
  • 驗證身分與工作階段,並執行對話或串流推論、RAG 增強與文字摘要
適用情境
當你希望助理直接呼叫託管的確定性合規、信任、身分或託管檢查而不必自行實作,或想在購買金鑰前先試用免費導引與示範呼叫時,值得加入。
執行需求
遠端 streamable HTTP 端點,無需本機安裝。免費導引工具不需要金鑰,但多數工具需要以 Authorization Bearer 權杖或 X-API-Key 標頭傳送的 API 金鑰,須向廠商商店購買。也可透過 x402 按次以微 USDC 付費呼叫。
安裝前請注意
多數工具需購買金鑰解鎖,呼叫可能按次以微 USDC 計費,費用會累積。託管工具會釋放資金並發起爭議,其他工具會登錄代理、SLA、事件與監督紀錄。原始碼未公開,README 說明部分連結的儲存庫為私有並回傳 404。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 Atomadic starter MCP (50 tools),將其新增到工作區。
  2. 為需要使用其工具的對話啟用該服務。

Web executable,透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。

交給代理程式安裝

把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式,或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南,先向你展示發布者、端點和登入方式,經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。

閱讀 https://sourceweft.com/mcp/io.github.atomadictech%2Faaaa-nexus/install.md,並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前,先告訴我它的發布者、端點和登入方式,等我確認。不要把 API 金鑰寫進命令或我會提交的檔案;問我金鑰應從哪裡取得。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。

{
  "mcpServers": {
    "atomadic-starter-50-tools": {
      "type": "http",
      "url": "https://mcp.atomadic.tech/mcp/public"
    }
  }
}

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:

bash
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.

The catalog also mirrors at https://aaaa-nexus.atomadictech.workers.dev/.well-known/mcp.json.

Want fewer tools? Use the starter door

https://mcp.atomadic.tech/mcp/public exposes a curated 50 of the 68 tools on the main endpoint. Same deterministic calls, same auth rules, smaller surface. GET it for a self-describing index. Start there if 68 is a lot to read.

The 30-second proof (no key, no account)

The full catalog lives at https://mcp.atomadic.tech (68 tools; a key unlocks your plan's set). Three tools need no key at all. Run this pair and you have seen the thing work:

bash
# 1. orientation: install, connect, what you are entitled tocurl -s -X POST https://mcp.atomadic.tech \  -H 'content-type: application/json' \  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"murmuration","arguments":{}}}'
# 2. the proof: one function emitted into 12 languages from one sourcecurl -s -X POST https://mcp.atomadic.tech \  -H 'content-type: application/json' \  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"render_polyglot_all_pure","arguments":{"atom_name":"compute_sum_pure"}}}'

That second call returns ok: true with Python, JavaScript, TypeScript, Rust, Go, C#, Kotlin, Swift, Ruby, Lua, Elixir and Haskell, all from a single node IR.

The baked demo atoms are compute_sum_pure, keep_positive_pure, double_each_pure, count_words_pure. Call the polyglot tools with an empty argument and they answer, honestly, novel_input_not_baked — that is a boundary, not a bug: arbitrary input is what the SDK is for.

Keys are sold at https://atomadic.tech/store. A key is emailed after checkout, and X-API-Key or Authorization: Bearer ato_<key> is the whole auth story.

How calls get paid

tools/call returns an invocation contract: an HTTPS URL, a method, and a price in micro-USDC. Pay it with either

  • x402 — complete payment per PAYMENT-REQUIRED on the 402 from the target route, or from /v1/payments/fee, sending an x402 PAYMENT-SIGNATURE header; or
  • X-API-Key — if you already hold a key.

No account is required for x402. Pay per call.

Elsewhere

  • https://atomadic.tech — main site, docs, pricing, proof page
  • https://aaaa-nexus.atomadictech.workers.dev — the server itself

On missing links

Some older pages and package metadata point at GitHub repositories that are private. Those links 404 for anyone outside the account. Nothing was deleted; the repositories are closed. If you followed a broken link to get here, this README is the correction: use the live endpoint above rather than the dead source link.

Registry listing

Official MCP Registry entry: io.github.atomadictech/aaaa-nexus (latest 0.5.2, remote https://mcp.atomadic.tech/mcp/public), with repository pointing at this repo. The older io.github.atomadictech/atomadic-forge entries are marked deprecated: that endpoint returns 404 and its package is retired.

https://registry.modelcontextprotocol.io/v0/servers?search=atomadic

Built by Thomas Ralph Colvin IV and Jessica Mary Colvin.

來源:README.md,提交 bda9797

工具

50

murmuration

Murmuration

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 — re-call anytime to re-check.

輸入參數agent_id

agents_contradiction

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.

輸入參數statements必填

agents_semantic_diff

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.

輸入參數contextoutput_a必填output_b必填

agents_register

Agents Register

[agents] Register an agent

輸入參數agent_id必填capabilities必填

agents_reputation

Agents Reputation

[agents] Agent reputation and trust score

輸入參數agent_id必填

agents_plan

Agents Plan

[agents] Autonomous agent goal planning

輸入參數goal必填constraints

reputation_record

Reputation Record

[agents] Record agent interaction

輸入參數outcome必填agent_id必填counterparty_idinteraction_type必填

sla_register

Sla Register

[agents] Register an SLA

輸入參數terms必填consumer_id必填provider_id必填

sla_report

Sla Report

[agents] Report SLA violation

輸入參數sla_id必填descriptionviolation_type必填

swarm_inbox

Swarm Inbox

[agents] Poll swarm inbox

無需輸入參數。

swarm_relay

Swarm Relay

[agents] Relay message through swarm

輸入參數ttlmessage必填target_agent_id必填

compliance_check

Compliance Check

[security] GDPR/CCPA compliance gate

輸入參數data_type必填operation必填jurisdiction必填

compliance_eu_ai_act

Compliance Eu Ai Act

[security] EU AI Act compliance certificate

輸入參數system_id必填risk_category必填intended_purpose必填

compliance_explain

Compliance Explain

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

輸入參數model_id必填decision_id必填subject_context必填

compliance_fairness

Compliance Fairness

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

輸入參數metric必填model_id必填dataset_id必填protected_attrs必填

compliance_lineage

Compliance Lineage

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

輸入參數model_id必填dataset_stages必填

compliance_transparency

Compliance Transparency

[security] Transparency report

輸入參數period必填include必填system_id必填

compliance_incident

Compliance Incident

[security] Report compliance incident

輸入參數severity必填system_id必填description必填affected_users必填

compliance_incidents

Compliance Incidents

[security] Incident registry

輸入參數system_id必填

compliance_oversight

Compliance Oversight

[security] Log oversight event

輸入參數notes必填decision必填system_id必填reviewer_id必填

security_zero_day

Security Zero Day

[security] Zero-day threat assessment

輸入參數target必填context

threat_score

Threat Score

[security] Multi-vector threat score

輸入參數contextagent_id必填

defi_optimize

Defi Optimize

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

輸入參數position必填protocol必填risk_tolerance必填

defi_risk_score

Defi Risk Score

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

輸入參數position必填protocol必填market_conditions必填

defi_yield_optimize

Defi Yield Optimize

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

輸入參數principal_usd必填risk_tolerance必填

escrow_create

Escrow Create

[defi] Create agent-to-agent escrow

輸入參數amount必填currency必填payee_id必填payer_id必填conditions

escrow_release

Escrow Release

[defi] Release escrowed funds

輸入參數escrow_id必填

escrow_dispute

Escrow Dispute

[defi] Open escrow dispute

輸入參數reason必填evidenceescrow_id必填

forge_badge

Forge Badge

[fuse] Model quality badge

輸入參數model必填

forge_leaderboard

Forge Leaderboard

[fuse] Model leaderboard

無需輸入參數。

forge_verify

Forge Verify

[fuse] Verify a model submission

輸入參數model_id必填verification_type

forge_delta_submit

Forge Delta Submit

[fuse] Submit a model delta

輸入參數delta必填model_id必填description

identity_verify

Identity Verify

[identity] Topological identity verification

輸入參數agent_id必填

identity_sybil_check

Identity Sybil Check

[identity] Sybil-resistance check

無需輸入參數。

ratchet_register

Ratchet Register

[identity] RatchetGate — register session

輸入參數session_id必填

ratchet_advance

Ratchet Advance

[identity] RatchetGate — advance session

輸入參數session_id必填

inference

Inference

[inference] Chat inference with hallucination guard

輸入參數messages必填

inference_stream

Inference Stream

[inference] Streaming inference with chain-of-thought

輸入參數messages必填

rag_augment

Rag Augment

[inference] Trusted RAG augmentation

無需輸入參數。

text_summarize

Text Summarize

[inference] Text summarization

輸入參數text必填

rng_quantum

Rng Quantum

[inference] Quantum random number with proof

無需輸入參數。

rng_verify

Rng Verify

[inference] Verify current epoch random value

無需輸入參數。

trust_score

Trust Score

[trust] Bounded trust score

輸入參數agent_id必填

drift_check

Drift Check

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

輸入參數model_id必填current_window必填reference_window必填

drift_certificate

Drift Certificate

[trust] Drift certificate

輸入參數id必填

oracle_hallucination

Oracle Hallucination

[trust] Hallucination oracle

輸入參數claim必填context

audit_work_claim

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. 3 free calls/day, then entitlement 'ledger'.

輸入參數inputs必填claimed_digest

verify_settlement

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'.

輸入參數inputs必填claimed_digest

detect_contradiction

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/day, then entitlement 'ledger'.

輸入參數claims必填

render_polyglot_all_pure

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.

輸入參數atom_name必填source_text必填

版本歷史

2
  1. v0.5.2最新Oct 29, 2026
  2. v0.4.0Sep 16, 2026

評分

已滿足適用檢查項的 90%

該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通必需25 分
  • 已宣告認證方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分

    請在儲存庫或註冊中心條目中宣告授權條款。

其他 Atomadic 整合