BankWeave

Name: BankWeave
Author: LSUDOKO

作者：LSUDOKO版本 3.0.1i★ –i电脑或云端

注册名称io.github.LSUDOKO/bankweave

安装

概览

作者提供的说明（英文原文）

Read-only FalkorDB tools for cross-bank AML investigations: rings, money paths, echo-proof claims.

PopNx MCP 的三大支柱

这三项永远不会合并成一个分数：意见、测量和人气是不同的东西。

用户评价i暂无评价评价功能将在后续阶段与账户功能一起开放
我们的基础检查i2 项中通过 1 项依据公开信息检查；尚未扫描代码
使用情况i尚未统计★ –

它能帮你做什么

我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。

连接方式: 电脑或云端
传输方式: stdio, Streamable HTTP
分类: 其他

安装

选择你使用的 AI 应用，然后复制设置。

安装之前

基础检查：2 项中通过 1 项未通过
在作者的服务器上运行：你向它提出的内容会经过他们的服务器。
注册中心没有列出需要填写的密钥。

你需要

在云端运行无需安装任何东西，通过互联网连接
Docker已安装 Docker

你用的是哪个应用？

ClaudeClaude CodeCursorVS CodeChatGPT

在上方选择你的应用。我们会记住它，下次按钮会显示“安装到 …”。

适用于网页版、桌面版和手机版 Claude：

打开 Settings → Connectors
点击 Add custom connector
粘贴下面的地址，然后点击 Add
如果要求登录，请登录作者的服务

https://bankweave-api.onrender.com/mcp

或者在你的电脑上运行

打开 Claude Desktop → Settings → Developer → Edit Config，把下面的内容粘贴到 claude_desktop_config.json

{
  "mcpServers": {
    "bankweave": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/lsudoko/bankweave-mcp:3.0.0"
      ]
    }
  }
}

在终端中运行这条命令

claude mcp add --transport http bankweave https://bankweave-api.onrender.com/mcp

或者在你的电脑上运行

claude mcp add bankweave -- docker run -i --rm ghcr.io/lsudoko/bankweave-mcp:3.0.0

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json，或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "bankweave": {
      "url": "https://bankweave-api.onrender.com/mcp"
    }
  }
}

或者在你的电脑上运行

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json，或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "bankweave": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/lsudoko/bankweave-mcp:3.0.0"
      ]
    }
  }
}

浏览器会先询问是否打开应用，然后应用会显示设置并请你确认。密钥的值会留空，由你自己填写。

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "bankweave": {
      "type": "http",
      "url": "https://bankweave-api.onrender.com/mcp"
    }
  }
}

或者在你的电脑上运行

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "bankweave": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/lsudoko/bankweave-mcp:3.0.0"
      ]
    }
  }
}

浏览器会先询问是否打开应用，然后应用会显示设置并请你确认。密钥的值会留空，由你自己填写。

ChatGPT 只能连接互联网上的服务器：

打开网页版 ChatGPT → Settings → Connectors (Apps)
点击 Create 或 Add（部分套餐需要先在 Advanced 中开启 Developer mode）
粘贴下面的地址，起个名字，然后保存
如果要求登录，请登录作者的服务

https://bankweave-api.onrender.com/mcp

菜单名称可能因套餐和版本不同而略有差异。

你的选择通过 Cookie 保存在这台设备上，无需账户。

这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同，请以 README 为准。

它能做什么

我们上次检查时服务器没有响应，所以无法读取它的工具。

基础检查i

我们检查的是任何人都能看到的信息；我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日

✓ 通过
作者与仓库一致
注册名称已通过 GitHub 账户 LSUDOKO 验证，该账户拥有此仓库
… 尚未检查
仍在维护
我们还没有从 GitHub 读取这个仓库
… 尚未检查
有许可证
我们还没有从 GitHub 读取这个仓库
✕ 未通过
端点有响应
我们检查时，6 秒内没有响应
– 不适用
使用 OAuth 登录
我们检查时，6 秒内没有响应
… 尚未检查
工具描述中没有隐藏指令
尚未检查；我们会在检查端点时读取工具列表

与“其他”中的其他服务器比较i

星标数多于此分类中其他 18386 个服务器的 0%
发布到注册中心的时间比此分类中其他 18386 个服务器的 68% 更新
基础检查结果等于或优于此分类中其他 18386 个服务器的 23%

与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时，星标数才计入。

常见问题

BankWeave 安全吗？

它在 2 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前，请先看看这个服务器要求哪些权限，并使用只具备所需权限的密钥。

它需要哪些密钥或代码？

注册中心没有列出必需的密钥，不过有些服务器会在首次使用时要求你登录。

它在我的电脑上运行还是在云端？

两种方式都可以：安装在你的电脑上，或者连接云端版本。

它还在维护吗？

GitHub 仓库最后更新于–，最新版本（3.0.1）于20天前进入注册中心。

哪些应用可以使用它？

支持 MCP 的应用，例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接，所以接受远程 MCP 服务器的应用也可以使用，例如开启开发者模式的 ChatGPT。

有什么可以替代它？

同一分类中有：Spec Driven Development, Remote Desktop Commander, Tethered Trust

替代选择

同一分类中其他作者的服务器，按星标数排序。

如何看卡片：★ GitHub 星标i三大支柱：评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i

Spec Driven Development

作者：formulahendry

MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.

在你的电脑上运行★ 440MIT14个月前更新

安装

Remote Desktop Commander

作者：desktop-commander

Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.

在云端运行★ 211前天更新

安装

Tethered Trust

作者：tetheredtrust.com

Fit, compatibility and recall checks for AI agents before they buy: verdicts, evidence, receipts.

在云端运行今天更新

安装

作者的 README

打开 README

按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。

BankWeave

Many banks, one ring, every claim proven.

BankWeave is a team of AI agents that investigates money-laundering alerts across several banks, with FalkorDB as its shared brain. It clears innocent alerts fast, exposes laundering rings that no single bank can see, and backs every conclusion with a graph path you can click through. It also knows how many *independent* reasons it has for each belief, notices when one of them dies, and refuses to file a report it can't prove.

Tracks: 01 Agents That Act on Connected Data (primary), 02 Agent Memory and Coordination, Best blog post
Live deployment: https://bankweave.vercel.app (UI on Vercel) · API https://bankweave-api.onrender.com/docs (Render) · graph: FalkorDB Cloud (ap-south-1)
MCP Registry: published as io.github.LSUDOKO/bankweave (v3.0.0, streamable HTTP)
Use it from Claude: claude mcp add --transport http bankweave https://bankweave-api.onrender.com/mcp, or as a stdio package: claude mcp add bankweave -- docker run -i --rm ghcr.io/lsudoko/bankweave-mcp (17 read-only FalkorDB tools, each returning the Cypher it ran; docs/MCP.md)
Proof in FalkorDB's own browser: screenshots and a recording of the ring, PageRank, card taps, SAR provenance and vector search running on our Cloud instance
Understand it in five minutes: the 3D explainer (eight scenes, each one FalkorDB call) and the graph explorer (twelve presets, run live)
Demo video: watch on YouTube, on the landing page or download the MP4 (narrated, captions burned in): a whiteboard explainer of the problem, the problem film's close, then the live product on FalkorDB Cloud, the hardware edge, MCP and FalkorDB's own browser. Build it yourself: video/build.py --long
Blog post: docs/BLOG.md

Running now on real data. The production instance on FalkorDB Cloud holds IBM's labelled AML transactions for 6 banks: 3,024 accounts, 24,990 transfers, 36 graphs (6 bank graphs, the consortium graph with 10,178 nodes and 37,448 edges, 27 case graphs, agent memory and the typology knowledge base). The agents have triaged 1,208 alerts: 1,113 cleared with a stated reason and 95 escalated into cases, each case with claims, evidence, the Cypher behind every piece of evidence, a provenance-gated report and a signed Merkle seal.

BankWeave is an independent project built for FalkorDB Graph Hacks. It is not a FalkorDB product and is unrelated to FalkorDB's QueryWeaver or Weights & Biases Weave.

The problem

Alert fatigue. Rule-based transaction monitoring flags single signals without context; industry sources put 85-95% of alerts as false positives.
Fragmented visibility. A bank only sees transactions that touch it. Laundering schemes spread across institutions, so each bank's slice looks innocent.
Unprovable reports. A suspicious activity report (SAR) has to justify itself, and today that justification is assembled by hand.

What BankWeave does

Loads IBM's public AMLworld laundering dataset into one isolated FalkorDB graph per bank plus a consortium graph that links them with salted, hashed account ids.
A team of agents (Orchestrator, Triage, Investigator, Skeptic, Reporter with a separate Auditor, Archivist) works the alert queue through a task blackboard stored in the graph, calling native FalkorDB algorithms as tools (also served over MCP).
Innocent alerts are cleared in milliseconds with a reason; suspicious ones become cases, each in its own case_ graph, where claims are linked to evidence and the query that produced it.
A trust layer built from three 2026 papers keeps the agents honest (below).
Humans approve every action. Everything is measured against IBM's labels on a scoreboard.

What's new in 3.0

Area	What you can do	Where
Console	A sidebar listing all fourteen pages with live counts, Ctrl K search over pages and cases, and one card system across every page: stat cards from live queries, status chips, rounded charts	every page
Dashboard	Open cases, alerts decided, independent claims, actions awaiting a human; claim support for the latest cases; the next approval; the agent team; an alert-triage gauge; FalkorDB Cloud latency, graphs and memory	`/command`
Graph explorer	A FalkorDB-browser-style page: pick a graph, write Cypher (`GRAPH.RO_QUERY`, writes rejected by the engine), see an Obsidian-style force graph with labels, hover neighbourhoods and a property panel. Twelve presets, each a FalkorDB feature on our data: cross-bank ring, `algo.pageRank`, `algo.labelPropagation`, `algo.wcc`, `algo.SPpaths`, Double Metaphone full-text, HNSW vector KNN, shared identity, the agents' alert-to-claim chain, a SAR's provenance, the typology KB, learned procedures	`/explorer`
3D explainer	Scroll through eight scenes on a three.js stage: six bank slices and per-graph ACL, hashed ids stitching the consortium, `algo.wcc` clearing an alert, the ring one bank sees 37.5% of, PageRank hub and `algo.SPpaths`, six paths collapsing to four origins, a KYC cascade, a Merkle root signed by a hardware key. Every number comes from the API; each scene shows the Cypher it ran	`/explainer`
Hosted MCP	`claude mcp add --transport http bankweave https://bankweave-api.onrender.com/mcp`: 17 read-only tools (cases, echo-proof claims, money paths, rings, hubs, components, vector and phonetic search, triage, typologies, guarded Cypher), 3 prompts, 2 resources	`/mcp`, docs/MCP.md
Landing	Centred hero over glossy blue bubbles with word-by-word type, the live console rising out of it, a "what's new" bento, the evaluation shown as highlighted screenshots of the live page, the real-tool recordings and the FalkorDB Browser proof	`/`
Real tools, recorded	The ESP32 sketch running in Wokwi posts a real card tap to the live API, and EMVLab decodes the field 55 it printed	Hardware Lab, below

Landing	Dashboard

Graph explorer: cross-bank ring on the consortium graph	3D explainer: the ring as one bank sees it

Graph explorer: `algo.pageRank` hubs (hover focuses one community)	3D explainer: six bank slices, each sees only its own graph

Alert queue	Case workspace

Approvals	Hardware Lab

Proof in the official FalkorDB Browser

The same data the agents use, opened in FalkorDB's own browser (browser.falkordb.com, connected to our FalkorDB Cloud instance: FalkorDB v4.22, 36 graphs, about 81 MB). Every query below is read-only and was typed into the browser; the timings are the engine's own RT readout. The connection bar with the instance hostname is cropped out.

Query	What it proves
Cross-bank ring: the case spanning the most banks, its accounts, banks and transfers (12 nodes, 166 edges, 1.85 ms)	one Cypher query sees the whole ring that no single bank graph contains
`CALL algo.pageRank('Account','TRANSFERS')` top 12 (3.55 ms)	the hub detection the Investigator calls is native GraphBLAS, not Python
`(:Account)-[:CARD_TXN]->(:Terminal)` (5 nodes, 29 edges, 0.41 ms)	the POS/ATM taps, including the one sent from the Wokwi ESP32, are real edges in the consortium graph; the Terminal node carries `POS-MUM-…`, Mumbai, GPS 19.076, 72.8777
`(:SAR)-[:CONTAINS]->(:Sentence)-[:CITES]->(:Claim)-[:SUPPORTED_BY]->(:Evidence)` in `case_C-70431130` (0.61 ms)	every report sentence cites a claim, and every claim rests on evidence: the provenance gate is a graph pattern
`db.idx.vector.queryNodes('Account','behavior_vec',10, …)` (0.36 ms)	HNSW vector search finds behaviour twins at five different banks
Accounts at different banks sharing a `Device` or `IP` (291 nodes, 220 edges, 2.79 ms)	the identity links the consortium uses to join bank slices

Cross-bank ring	`algo.pageRank`

Card taps into terminals (the Wokwi tap included)	The Mumbai POS node and its properties

SAR → sentence → claim → evidence	HNSW vector KNN

Shared devices and IPs across banks	FalkorDB Cloud console: the instance (Free tier, ap-south-1, v4.22.0, healthy)

Hosted MCP, called live (no install, plain HTTP):

$ curl -s -X POST https://bankweave-api.onrender.com/mcp -H 'Content-Type: application/json' \

-H 'Accept: application/json, text/event-stream' \ -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' 17 tools: engine_overview, list_graphs, graph_schema, list_cases, get_case, explain_claim, account_network, trace_money_path, find_rings, top_hubs, connected_components, similar_accounts, search_customers, list_alerts, triage_alert, typology_guidance, run_readonly_cypher

$ ... "method":"tools/call","params":{"name":"top_hubs","arguments":{}} {"graph": "consortium", "result": [{"account": "e8ccd1faa8d3a4ba", "bank": "001", "score": 0.0027...}, ...

The top account matches row 1 of the PageRank query in the FalkorDB Browser above.

What's new in 2.0

Area	What you can do	Where

| Command Center | Live alerts, open cases, independently supported claims, evidence awaiting review, FalkorDB healt

只显示了开头部分；其余内容请在仓库中阅读

BankWeave

概览

PopNx MCP 的三大支柱

它能帮你做什么

安装

安装之前

或者在你的电脑上运行

或者在你的电脑上运行

或者在你的电脑上运行

或者在你的电脑上运行

它能做什么

基础检查i

作者与仓库一致

仍在维护

有许可证

端点有响应

使用 OAuth 登录

工具描述中没有隐藏指令

与“其他”中的其他服务器比较i

常见问题

替代选择

Spec Driven Development

Remote Desktop Commander

Tethered Trust

作者的 README

BankWeave

The problem

What BankWeave does

What's new in 3.0

Proof in the official FalkorDB Browser

What's new in 2.0