
Fidacy
com.fidacy由 fidacy.com 發布v0.2.7更新於 Oct 24, 2026
Payment firewall for AI agents: a signed, verifiable verdict on every money-moving action.
概覽
本機 MCP 伺服器,在資金移動前依簽章授權書攔截 AI 代理的付款動作,回傳允許/拒絕結論、短期授權憑證與稽核證明。
- 可以做什麼
- 依簽章授權書請求並授權付款,回傳附短期 Ed25519 憑證的允許結果,或附違規規則的拒絕結果
- 讀取授權書信封與用於驗證的 Fidacy 公鑰
- 依決策 ID 取得雜湊鏈稽核證明
- 向 Fidacy 引擎請求對擬執行動作(如付款、發訊息、語音通話、文件聲明)的簽章信任結論
- 適用情境
- 當代理具備動用資金的能力,而你希望在執行器之前加一道確定、可驗證的關卡,而非依賴提示詞層級的防護,以便在資金流出前攔下錯誤收款方、虛高金額或重複請款時使用。
- 執行需求
- 透過 npx 或全域 npm 安裝,在本機以 stdio 執行(需要 Node.js)。可離線使用,並在本機設定檔中設定可信收款方與額度上限。選用的 assess_action 結論需要具備 assess:write 權限的 FIDACY_ENGINE_API_KEY;接上正式核心需設定 FIDACY_MODE=http 並實作三個 HTTP 端點。
安裝
在 SourceWeft 中
- 開啟 儀表板中的 Fidacy,將其新增到工作區。
- 為需要使用其工具的對話啟用該服務。
Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。
其他 MCP 客戶端
參照 儲存庫 中的啟動說明。
README
@fidacy/mcp
[Fidacy blocks a BEC lookalike-payee payment, then allows the legit one with a signed Ed25519 grant]
The action firewall for AI agents. A drop-in MCP server that gates payment actions against a cryptographically signed mandate before money can move. Non-custodial: Fidacy authorizes and proves, it never holds funds.
Install once, works in any MCP-compatible agent: Claude Code, Claude Desktop, Hermes, OpenClaw, and anything else that speaks MCP.
[npm] [license] Works with: Claude Code · Claude Desktop · OpenClaw · Hermes · Brex CrabTrap
Your agent could be paying scammers right now. Prompt-injected into the wrong payee, an inflated amount, or the same invoice twice — and your logs aren't evidence. Fidacy blocks it before money moves, and hands back a signed verdict anyone can verify against public keys. You don't trust us — you check the signature.
Quick start (free, local-first, no account)
Runs on your machine, offline, deny-by-default. Add trusted payees + caps in
~/.fidacy/config.json. Verify any verdict yourself against the public keys at
/.well-known/jwks.json.
Why
An agent can hallucinate or be prompt-injected into a payment: wrong payee,
wrong amount, fabricated invoice. Prompt-level guardrails are probabilistic and
bypassable. @fidacy/mcp is a deterministic gate between the agent's intent and
the executor: the action is dead on arrival unless it validates against a signed
mandate, and every decision lands in an immutable hash-chained audit trail.
Enforcement model
- Register
@fidacy/mcpas the agent's only payment-capable tool. Do not give the agent a raw payment tool. Tool inventory is the runtime firewall. - The agent calls
request_payment. Fidacy checks it against the mandate (payee allowlist, per-tx cap, total cap, currency, time window, revocation). - ALLOW returns a short-lived Ed25519 grant. DENY returns no grant and the violated rule. The downstream executor MUST require the grant, so a denied action cannot proceed.
- Every decision is appended to a hash-chained log.
get_audit_proofreturns the portable, verifiable proof.
One install, two backends
@fidacy/mcp ships two complementary capabilities in a single install:
- Verdict layer (advisory):
assess_actioncalls the live Fidacy engine and returns a signed trust verdict. It moves no money; it returns a judgment whose proof (riskPayloadJws+signingKeyId) is verifiable by anyone via@fidacy/verifyagainst the engine JWKS at/.well-known/jwks.json. - Payment firewall (enforcement):
request_payment/verify_mandate/get_audit_proofgate and prove a payment against a signed mandate through the core, returning short-lived Ed25519 grants.
Mental model: assess_action -> engine (signed verdict);
request_payment and friends -> core (payment firewall).
Tools
assess_action
Returns a signed Fidacy trust verdict from the live engine for a proposed
action. The signed proof is riskPayloadJws + signingKeyId, verifiable by
anyone via @fidacy/verify against {engineUrl}/.well-known/jwks.json.
Inputs:
kind(optional, defaultap2_payment): one ofap2_payment,message_send,voice_call,custom,claim_document.mandate(required): the action/mandate object for thatkind.mandateType,spendingMandate,idempotencyKey,a2a.task_id(optional).
Environment:
The server boots without FIDACY_ENGINE_API_KEY; the tool is always registered.
Only calling assess_action without the key returns a helpful error telling
you to set it. The key is never logged, echoed, or attached to any error.
Install
Claude Code
Claude Desktop (claude_desktop_config.json)
Hermes (config.yaml)
OpenClaw
Add the same server via the Tools panel, or the mcpServers block in your
agent config. Any MCP-compatible host uses the same command.
Wiring the real core (production)
The MCP layer talks to your core through one interface (FidacyCore). Your
repository stays private. Set FIDACY_MODE=http and implement three endpoints:
POST /v1/mandate/get->MandatePOST /v1/decide->Decision(runs your Ed25519/AP2 verification + audit append)POST /v1/audit/proof->AuditProof
No change to the MCP layer is needed.
Dev
來源:README.md,提交 ce42c70
工具
0版本歷史
1- v0.2.7最新Sep 16, 2026
評分
已滿足適用檢查項的 50%
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25 分
請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。
- 已宣告認證方式10 分
請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。
- 已列出工具15 分
請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。
- 描述必需15 分
- 圖示10 分
- 文件15 分
- 授權條款10 分