Give your agents live, governed intelligence from the first call.
MCP-native. API-first. Zero Trust by default.
$ chordian connect --tool cursor
✓ Chordian connected · Zero Trust policies active
Works with Claude Code, Cursor, Codex, Gemini CLI, Windsurf and any MCP client.
What your agent can call
On every call: structured JSON, sources, confidence and conflict flags.
| Solution | Skills |
|---|---|
| Capital Markets | Investor origination · Buyer mapping · Fund intelligence · Meeting prep |
| Real Estate | Market intelligence · Approval tracking · Counterparty checks · Off-market sourcing |
| Sales & Distribution | Leads · Account research · Signal tracking · Distribution partner search |
| Consulting & Research | Research · Competitor checks · Product intelligence · R&D scanning |
| Due Diligence & KYC/KYB | Diligence · KYB · KYC · Filing monitoring |
| Cyber & Leak Monitoring | Monitoring · Leak detection · Executive protection · Threat briefs |
Built for production
Context-efficient
5,016 median context tokens per query on our benchmark.
Permission-scoped
Every result filtered by ABAC for the calling agent or user.
Auditable
Every agent call in a tamper-evident log.
Deploy anywhere
The same API in Chordian Cloud, your cloud, on-prem or air-gapped.
Let your agent set Chordian up itself.
Paste this into Claude Code, Cursor, Codex or any coding agent:
| Function | Use it when your agent needs… |
|---|---|
| Web Intelligence | a sourced answer from the live web (Fast or Deep) |
| People & Companies | verified, enriched entity profiles or lists |
| Agentic Crawler | structured data extracted from specific sites, registries or documents |
| Scheduled Search | to watch something and react when it changes (webhooks) |
| Memory Search | permission-scoped answers from your organization's own knowledge |
Your agent reads agents.md
Capabilities, auth, tool names, examples.
It installs the Chordian MCP server
And requests your API key.
It runs a test call
And confirms Zero Trust policies are active.
Prefer manual setup? MCP Server → API Reference → agents.md →
Chordian MCP Server
Five tools. Every MCP client. Governed by default.
| Tool | Does |
|---|---|
web_intelligence | question → sourced answer (depth: fast | deep, country) |
people_companies | query or list → verified profiles |
agentic_crawler | URLs or sources plus schema → structured dataset |
scheduled_search | query plus schedule → webhook alerts |
memory_search | question → permission-scoped internal answer |
Permissions: each MCP connection maps to an identity. ABAC policies decide what that agent can read, write and search, and every call is logged.
Full MCP docs ↗Chordian API
REST endpoints for all five functions plus Memory read and write. JSON in, JSON out, with sources, confidence and conflicts on every result.
- Authentication: API keys, SSO, BYOK for self-hosted deployments
- Async jobs and webhooks for Deep and Agentic Crawler runs
- The same API across Chordian Cloud, BYOC, on-prem and air-gapped