Cipher & Row

by cipherandrowhq

Not rated
GitHub

About

Verify US/Canada freight carriers and brokers: live FMCSA data, trust scores, registry search, and partner watchlist monitoring for AI agents.

Details

Author
cipherandrowhq
Categories
Other, Security, API

Setup

Install Cipher & Row in your MCP client (Claude Desktop, Cursor, Windsurf, and others).

Repository: https://github.com/cipherandrowhq/cipherandrow-mcp

Follow the installation instructions in the repository README, then restart your MCP client.

Cipher & Row - Freight Verification MCP

LLM-ready freight data for AI agents.Live FMCSA carrier & broker verification, multi-signal 0-100 trust scores, double-brokering and chameleon-carrier fraud signals, and a monitored partner watchlist - as nativeModel Context Protocoltools an autonomous dispatch or brokerage agent can call directly.

This repository is theopen, public interfacefor the Cipher & Row MCP server: the exact tool JSON Schemas, copy-paste client configs, and integration guides. Theverification engine, scoring model, and infrastructure are proprietary(seeNOTICE); this repo is the contract you build against, kept in sync with the live server.

Who this is for:engineers buildingAI agent carrier vetting, autonomous dispatch, brokerage automation, and freight-fraud-prevention systems - and security reviewers evaluating how the server resists prompt injection.

An AI agent in freight reads untrusted text all day and can be talked into fraud - a fake carrier, a spoofed broker, a double-brokering scam. The fix is to never let the modeldecidetrust from the conversation, but toverifyagainst authoritative records the attacker can't reach. This MCP server is thatair-gapped verification layer: you verify byDOT / MC / CAB numberagainst FMCSA, surety-bond, sanctions, and registry data, and get back structured evidence plus a canonical trust score - not a persuadable opinion. Full model:docs/PROMPT-INJECTION-DEFENSE.md.

claude mcp add --transport http cipher-row https://mcp.cipherandrow.com/mcp # then run /mcp in a session and choose Authenticate

Claude Desktop / stdio clients- pastemcp-config/claude-desktop.jsoninto yourclaude_desktop_config.json.

Cursor, Windsurf, others- seemcp-config/cursor-and-other-clients.md.

Raw HTTP- OAuth 2.1 + PKCE, or anAuthorization: Bearer cr_live_...key:

curl -sS https://mcp.cipherandrow.com/mcp \ -H "Content-Type: application/json" \ -H "Authorization: Bearer cr_live_YOUR_KEY" \ -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{ "name":"cr_verify_carrier","arguments":{"dot_number":"53467"}}}'

Full manifest:schemas/tools.json· Human reference:docs/TOOLS.md.

The surface is verification + a watchlist. No tool posts a load, moves money, or messages a counterparty- so even a hijacked agent's blast radius is bounded by design.

├── schemas/ JSON Schemas (draft 2020-12) for all 8 tools + tools.json manifest ├── mcp-config/ Copy-paste client configs (Claude Desktop/Code, Cursor, generic) ├── docs/ │ ├── PROMPT-INJECTION-DEFENSE.md Air-gapped verification: how the server resists injection │ ├── API-TO-MCP.md Append-only audit trail + multi-signal scores over MCP │ ├── AUTHENTICATION.md OAuth 2.1 + PKCE and API-key auth │ ├── RATE-LIMITS.md Per-tier, per-category ceilings │ └── TOOLS.md Human-readable tool reference ├── examples/ │ ├── vet-a-carrier.md End-to-end disambiguate → verify → decide → monitor │ └── sample-responses/ Illustrative response shapes ├── SECURITY.md Responsible disclosure + security posture ├── LICENSE MIT (this interface repo) └── NOTICE What is open here vs. what stays proprietary

- Disambiguatewithcr_search_registry(identity only, no score).
- Verifywithcr_verify_carrier/cr_verify_broker- profile, fraud signals, and a canonical0-100 trust scorewith a per-factor breakdown, computedserver-sidefrom independent regulated signals. The entity being scored has no channel to inflate its own score.
- Decidein your agent, on the evidence and verdict - not on the conversation.
- Monitorwithcr_add_partnerso authority, insurance, and bond changes reach you between deals. Every verification call is written to aserver-side audit log.

- United States- live FMCSA authority, insurance (BIPD/cargo/bond), safety and inspection history, surety bonds (BMC-84/85), sanctions screening.
- Canada- federal + cross-border FMCSA on every lookup; provincial depth varies by province. Canadianbrokerverification is available over MCP (cab_id); Canadiancarrierintake is a REST write, not an MCP tool.

- Air-gapped verification- trust decisions are grounded in out-of-band authoritative data, not in the agent's persuadable context.
- Typed output is structured evidence; third-party free-text fields are data to reason over, never commands to execute.
- OAuth 2.1 + PKCEwith scoped, server-enforced role gating andper-key rate limits.
- EIN excludedfrom the agent door; inputs sanitized; payloads capped.
- Audit logof every verification call (which tool, which account, when, outcome).

The contents of this repository - schemas, configs, and documentation - are released under theMIT Licenseso you can freely build against, redistribute, and adapt the interface. The Cipher & Row verification server, scoring model, and engine arenotopen source; seeNOTICE.

Setup + live examples:https://www.cipherandrow.com/mcp/setup·Partnerships / review keys:partnerships@cipherandrow.com

Keywords: MCP freight verification · LLM-ready freight data · AI agent carrier vetting · FMCSA carrier lookup API · broker verification · double-brokering fraud detection · chameleon carrier · freight trust score · Model Context Protocol logistics · autonomous dispatch agent.

EU Corporate Sustainability Reporting Directive compliance — ESRS mapping, double materiality, ESG data collection by MEOK AI Labs

MCP-сервер для проверки контрагентов через egrul.nalog.ru: получение выписки ЕГРЮЛ/ЕГРИП по ИНН/ОГРН.

CLI and MCP server for the UK Companies House API — company search, profiles, officers, filings, ownership, and due diligence

Company lookup, LEI search, SEC filings, and financials for AI agents. 6 tools, free, no API key.

All-in-one bundle: EU VAT validation, Dutch CBS statistics, and GDPR compliance tools — 19 tools for EU businesses in a single MCP server.

Ask natural language questions about your SafetyCulture data using the SafetyCulture API.

Provides basic enterprise information services, including business registration, company profiles, shareholders, and key personnel.

Provides comprehensive enterprise risk analysis, including violation records, mortgage information, business anomalies, and judicial cases.

Provides comprehensive information queries for enterprise qualification certificates, including honors, administrative licenses, and profiles.

A Model Context Protocol (MCP) server that provides comprehensive access to the ALMA (Atacama Large Millimeter/submillimeter Array) archive through a clean, extensible architecture.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.