FMCSA Carrier Vetting

io.github.santex85由 GitHub 账号 santex85 发布v0.1.0更新于 Nov 19, 2026

Vet US motor carriers and brokers from FMCSA open data: authority, insurance, OOS, safety.

发布方自己的服务:Carrier

已验证STDIO仅桌面数据与分析商业与电商

概览

AI 生成的概览

基于 FMCSA 公开数据核查美国汽车承运人和经纪人:查询、名称搜索、资质与保险合规检查,以及启发式欺诈风险标记。

可以做什么
  • 按 USDOT 或 MC/FF/MX 编号查询承运人档案,含状态、备案、地址、车队规模与安全评级
  • 按法定名称或 DBA 名称搜索承运人,可按州、城市、仅活跃、仅营利性运输筛选
  • 查看完整资质与保险合规情况:各类资质及状态、在册的 BIPD/货物险/保证金、授予日期、取消记录、吊销与停运令,以及 24 个月检查与事故计数
  • 运行启发式欺诈/风险筛查,返回按严重程度编码的标记、评分和风险等级
适用场景
适合货运经纪人、托运人和第三方物流在派单前确认承运人身份、是否持有有效资质与保险,以及是否存在欺诈迹象。
运行要求
以本地 Python 进程运行(stdio 或 streamable HTTP);README 说明需创建虚拟环境并安装依赖,清单另提供 mcpb 包。无需 API 密钥或登录,直接查询 data.transportation.gov 上的 FMCSA 开放数据集。可选环境变量 HOST、PORT、MCP_TRANSPORT。需要网络访问。
安装前请注意
风险标记仅为启发式判断,并非欺诈认定,应拨打 FMCSA 列出的电话核实。数据可能比 SAFER/L&I 实时页面滞后一天以上,检查与事故数字为原始计数,并非官方 SMS 百分位。大量生产流量可能触发匿名 Socrata 限流。仅返回 FMCSA 公开的企业联系信息。

安装

在 SourceWeft 中

  1. 打开 控制台中的 FMCSA Carrier Vetting,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

FMCSA Carrier Vetting MCP server

An MCP server for freight brokers, shippers and 3PLs that need to vet US motor carriers and brokers before tendering a load. It answers "who is this carrier, are they allowed to haul, are they insured, and does anything look like fraud?" in one call. It uses FMCSA's own open data on data.transportation.gov (Socrata SODA API, no key, no login).

Tools

ToolWhat it does
lookup_carrierCompany profile by USDOT or MC/FF/MX number: legal/DBA name, USDOT status, entity class, dockets, address, business phone, power units, drivers, cargo types, MCS-150 date, safety rating, prior-revocation flag, and an operating-authority summary
search_carriersName search (legal or DBA, substring) with state, city, active-only and for-hire-only filters, up to 100 rows, largest fleets first
check_authorityFull compliance view: every authority (carrier/broker/HHG/passenger) with status, BIPD required vs on file, cargo/bond, authority grant dates and status history, insurance filings, cancellations, revocation/suspension orders, out-of-service orders, plus 24-month inspection and crash counts
red_flagsHeuristic fraud/risk screen with severity, evidence and a risk level (see below)

Example calls

jsonc
// lookup_carrier{"usdot": "80806"}{"mc_number": "MC-136818"}          // also "MC136818", "136818", "FF-51075"
// search_carriers{"name": "swift transportation", "state": "AZ", "active_only": true, "limit": 10}{"name": "express", "state": "TX", "city": "HOUSTON", "for_hire_only": true}
// check_authority{"usdot": "80806"}
// red_flags{"mc_number": "MC-1589648"}

Sample red_flags output for a carrier whose authority was granted the day before:

json
{"usdot": "6951245", "riskLevel": "high", "score": 80, "flags": [{"severity": "high", "code": "NEW_USDOT", "message": "USDOT number added 14 days ago"},           {"severity": "high", "code": "NEW_AUTHORITY", "message": "Oldest operating-authority grant on record is only 1 days old"}]}

Red-flag checks

CodeSeveritySignal
NOT_FOUND, USDOT_INACTIVEhighUSDOT unknown or not active
NO_ACTIVE_AUTHORITYhigh (low for private/intrastate entities)no active MC/FF/MX authority
NEW_AUTHORITY, NEW_USDOThigh < 90 days, medium < 180 daysfreshly created identity, typical of fraud
PENDING_REVOCATION, RECENT_INVOLUNTARY_SUSPENSIONhigh / mediumFMCSA orders not yet effective; suspensions in the last 12 months
OUT_OF_SERVICE / PAST_OUT_OF_SERVICEhigh / lowFMCSA OOS order in effect / rescinded
INSURANCE_MISSING, INSURANCE_CANCELLATION_PENDING, INSURANCE_CHURNhigh / high / mediumBIPD below required, cargo or bond missing, future cancellation, 2+ cancellations in 12 months
PRIOR_REVOCATIONhighFMCSA census prior-revocation flag (possible "chameleon" carrier)
SHARED_PHONE, SHARED_ADDRESShigh/medium/lowthe census phone or physical address appears on other USDOTs (downgraded for fleets of 100+ trucks, whose affiliates share numbers)
NAME_MISMATCH, ADDRESS_MISMATCHlowCompany Census vs Motus L&I record differ
STALE_MCS150low/mediumMCS-150 older than 2 / 3 years
UNSATISFACTORY_RATING, CONDITIONAL_RATINGhigh / mediumFMCSA safety rating
NO_POWER_UNITS, HIGH_OOS, NO_INSPECTIONSmedium / medium / lowcarrier authority with 0 trucks; many OOS violations; no inspections despite 5+ trucks

Score: high = 40, medium = 15, low = 5, capped at 100. Risk level is high if any high flag (or score ≥ 60), medium if score ≥ 15, otherwise low. These are heuristics, not a finding of fraud. Always confirm by calling the phone number FMCSA lists, not the one in an e-mail.

Data sources

All are FMCSA datasets on data.transportation.gov, read live per call (nothing is downloaded or cached except the per-dataset "last updated" timestamp, kept 1 hour):

DatasetIdUsed for
Company Census Fileaz4n-8mr2profile, search, shared phone/address
Motus Carrier - All With Historyinys-ebihoperating authority per type, BIPD/cargo/bond on file
Motus AuthHist - All With Historyyu5v-wbh6grant dates, status changes
Motus Insur - All With Historyc5y8-a4uzactive/pending insurance filings
Motus InsHist - All With History3uet-3z4iinsurance cancellations
Motus RevokeSuspend - All With Historywb4f-nekirevocation/suspension orders
Out Of Service Ordersp2mt-9igeOOS orders
Vehicle Inspection Filefx4q-ay7w24-month inspection/OOS counts
Crash Fileaayw-vxb324-month crash counts
AuthHist - All With History (legacy L&I)9mw4-x3tugrant dates before the Motus migration

Freshness. Every response has a dataFreshness list with each dataset's rowsUpdatedAt. FMCSA refreshes these daily (all were updated 2026-11-17 when this was built). Data can lag the live SAFER / L&I web pages by a day or more. Motus is FMCSA's current registration system. The legacy L&I files are still published, but the legacy authority history has no new actions after about May 2026, so this server reads authority and insurance from the Motus files and uses the legacy history only for older grant dates.

Terms / licence. The datasets carry no explicit licence field on data.transportation.gov. They are US federal government works published as open data, generally public domain in the US (17 U.S.C. § 105); this was not confirmed against a dataset-specific licence. No key is used. Socrata throttles anonymous clients, so for heavy production traffic register a free Socrata app token (not used here). The optional FMCSA QCMobile API needs a webKey via login.gov and is not used.

Privacy. Only what FMCSA publishes is returned, with no enrichment. The census file also contains company officer names, e-mail addresses and cell phones. This server deliberately leaves those out and returns only the business phone and addresses that the SAFER Company Snapshot shows.

Limits

  • 90 s timeout per tool call, 45 s per HTTP request, 2 retries on 429/5xx.
  • search_carriers: name of 3+ characters, at most 100 results (scans up to 500 matches to sort by fleet size).
  • History lists are capped (50 status changes, 50 insurance filings, 25 cancellations, 50 orders, 10 shared-contact examples).
  • Inspection/crash numbers are raw counts from FMCSA files, not the official SMS BASIC percentiles.
  • One SODA query per dataset per call; red_flags and check_authority make about 10 to 13 requests and usually take 5 to 12 s.

Run

bash
python3 -m venv .venv && .venv/bin/pip install -r requirements.txt.venv/bin/python src/server.py                         # streamable-http on 0.0.0.0:8080, endpoint /mcp, GET /healthPORT=8094 .venv/bin/python src/server.py               # other portMCP_TRANSPORT=stdio .venv/bin/python src/server.py     # stdio, for local MCP clients.venv/bin/python tests/live_test.py                    # live test of every tool

Environment variables: HOST (default 0.0.0.0), PORT (default 8080), MCP_TRANSPORT (streamable-http, stdio or sse; default streamable-http). No API key is needed.

mcpize.yaml is included for MCPize (runtime: python, startCommand.type: http).

Use with Claude Desktop or another MCP client

Local (stdio). Clone the repo, create the venv as above, then add this to claude_desktop_config.json (or your client's MCP config), with the paths changed to where you cloned it:

json
{  "mcpServers": {    "fmcsa-carrier": {      "command": "/path/to/fmcsa-carrier-mcp/.venv/bin/python",      "args": ["/path/to/fmcsa-carrier-mcp/src/server.py"],      "env": {"MCP_TRANSPORT": "stdio"}    }  }}

Remote (streamable HTTP). Run .venv/bin/python src/server.py on a server and point any client that supports remote MCP servers at http://<host>:8080/mcp. GET /health returns {"status": "healthy"}.

License

MIT, see LICENSE. The data returned by the tools stays under the terms of its sources (see above).

来源:README.md,提交 72d3f91

工具

0
SourceWeft 尚未在沙箱中启动过该软件包,因此其工具未知。

版本历史

1
  1. v0.1.0最新Nov 19, 2026

评分

已满足适用检查项的 50%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分

    请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。

  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。

  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分

其他 Carrier 集成