1,560 live data tools.
Loaded on demand.
1.7M+ requests/month 13K+ unique visitors/month
One MCP gateway URL gives your AI agent on-demand access to 419 sources — SEC, FDA, FRED, ClinicalTrials, USPTO, EPA, and 413+ more. Your agent loads only the ~20 tools it needs per session, so you don't burn context on definitions you'll never use. No stale training data. No rebuilding 50 integrations. No API key management.
npx pipeworx use weatherOr add any pack directly to Claude Desktop — no CLI required. Browse packs →
Why Pipeworx
Connector libraries built three billion-dollar businesses for the previous generation of tooling. We're doing it for AI agents — with the protocol, reliability, and edge-native architecture the agent era actually needs.
The connector library is the moat.
Airbyte ($1.8B, ~600 connectors) and Fivetran ($5.6B, ~700+) built billion-dollar businesses on one insight: developers don't want to write integrations one at a time. Plaid ($13.4B) proved the same model in a single vertical.
Pipeworx has parity on connector breadth — 1,560 tools across 419 live data sources, unified behind one protocol. The integration layer is being rebuilt for a new consumer: not data pipelines, but agents.
Full MCP, not just tool calling.
The MCP spec has three primitives — Tools, Resources, Prompts. Most MCP servers ship
only Tools. Resources let agents cite by URI for durable output
(pipeworx://edgar/company/{cik}/filings). Prompts encode optimal
multi-step workflows.
Pipeworx ships all three, plus seven gateway-native meta-tools
(ask_pipeworx, discover_tools, resolve_entity,
compare_entities, entity_profile, recent_changes,
remember/recall/forget) that collapse 5–15 agent
round-trips into one call. One entity_profile call on a ticker returns
filings + fundamentals + patents + news + LEI in a single round trip.
Reliability is the product.
Agent tool calls fail 25–30% of the time in production against typical MCP servers. Almost no one measures it. We instrument every call, classify errors, and ship telemetry-driven fixes.
Last week we cut our own error rate from 27.6% to 6.9% in one afternoon — every remaining error is a legitimate user/config issue, not a bug. Read how →
vs Composio / Pipedream MCP / Zapier MCP: we ship the full MCP protocol surface and target autonomous-agent traffic, not human-triggered workflows. vs Smithery / mcp.so / Glama: they're directories; we're one MCP server with 419 sources behind it. vs DIY: every team rebuilds the same integrations one source at a time. See full comparisons →
Featured packs
Twelve of 419 hosted packs · 1560 tools across the catalog · free tier · browse all 419 packs →
Edgar
EDGAR MCP — SEC EDGAR public APIs (free, no auth)
Fred
FRED MCP — Federal Reserve Economic Data (St. Louis Fed)
Openfda
OpenFDA MCP — wraps the openFDA API (free, no auth required)
Clinicaltrials
ClinicalTrials MCP — wraps ClinicalTrials.gov API v2 (free, no auth)
Attom
ATTOM MCP — Premium real estate data from ATTOM Data Solutions
Census
Census MCP — U.S. Census Bureau housing-relevant APIs.
Patents
Search US patents by keyword, retrieve patent details, and look up inventors via PatentsView
Epa Echo
EPA ECHO MCP — wraps EPA ECHO Web Services (free, no auth)
Samgov
SAM.gov MCP — Federal contract opportunities and entity registration data
Weather
Real-time weather conditions and multi-day forecasts from Open-Meteo
Wikipedia
Search and retrieve Wikipedia articles — summaries, section structure, and random article discovery
Crossref
Search and retrieve academic paper metadata — DOIs, authors, citations, and journals via Crossref
How it works
A real example: agent connects to Pipeworx, asks for Apple's full profile, gets SEC filings + fundamentals + patents + news + LEI back in one round trip — across five packs, in parallel, with citation URIs.
Add Pipeworx to your MCP client
Drop one JSON block into claude_desktop_config.json, Cursor settings, or
any MCP-aware client. No API keys, no SDK install.
{
"mcpServers": {
"pipeworx": {
"url": "https://gateway.pipeworx.io/mcp"
}
}
}Your agent makes one call instead of fifteen
User asks for a company brief. The agent calls one Pipeworx meta-tool —
entity_profile — which fans out across SEC EDGAR, SEC XBRL, USPTO, GDELT,
and GLEIF in parallel inside the gateway. No orchestration code, no sequential
round-trips burning context.
User: Give me a brief on Apple.↓Agent calls: entity_profile({ type: "company", value: "AAPL" })
Pipeworx returns five sources of structured data with citations
One JSON-RPC call returns SEC filings, fundamentals (Revenue / NetIncome / Cash from
XBRL), USPTO patents, recent news, and the LEI — each item carrying a
pipeworx:// URI for stable citation. Partial failures (e.g., a rate-limited
upstream) surface in sources_failed rather than breaking the whole call.
{
"ticker": "AAPL",
"cik": "0000320193",
"company_name": "Apple Inc.",
"lei": "HWUPKR0MPOU8FGXBT394",
"recent_filings": [
{ "form": "10-Q", "date": "2026-05-01",
"_pipeworx_uri": "pipeworx://edgar/company/0000320193/filings/..." },
{ "form": "8-K", "date": "2026-04-30", "_pipeworx_uri": "pipeworx://edgar/..." }
],
"fundamentals": {
"Revenues": { "value": 416161000000, "unit": "USD", "period": "2025-09-27" },
"NetIncomeLoss": { "value": 112010000000, "unit": "USD", "period": "2025-09-27" },
"CashAndCashEquivalentsAtCarryingValue":
{ "value": 29943000000, "unit": "USD", "period": "2024-09-28" }
},
"recent_news": [...],
"sources_used": ["edgar", "sec-xbrl", "gleif", "gdelt"],
"sources_failed": []
}Common questions
What is Pipeworx?
Pipeworx is an open MCP gateway that connects AI agents to live data from 419 sources — SEC EDGAR, FDA, FRED, the Census Bureau, USPTO, EPA, UN Comtrade, ClinicalTrials.gov, and many more. One gateway URL exposes 1,560 tools, and your agent loads only the ones it needs for the task at hand.
How do I connect my AI agent to Pipeworx?
Add a single MCP server entry pointing at https://gateway.pipeworx.io/mcp. The gateway speaks the full Model Context Protocol — Tools, Resources, and Prompts — so it works with Claude Desktop, Claude Code, Cursor, VS Code, ChatGPT, Windsurf, and any other MCP-compatible client without an SDK install.
What is MCP (Model Context Protocol)?
MCP is an open protocol for connecting AI agents to external tools and data. It standardizes how an agent discovers, calls, and cites data sources. The spec defines three primitives — Tools (function calls), Resources (citable URIs), and Prompts (workflows). Pipeworx ships all three, plus seven gateway-native meta-tools that collapse multiple round-trips into one.
Do I need to manage API keys for each data source?
No. The Pipeworx gateway handles authentication for public sources (FRED, SEC EDGAR, FDA, Census, BLS, EPA, ClinicalTrials, USPTO, and many more) without any keys at all. For premium sources like ATTOM, Altos, or Alpha Vantage you can bring your own key, or use metered Pipeworx access on a paid plan.
How is Pipeworx different from Composio, Smithery, or Zapier MCP?
Composio, Pipedream, and Zapier MCP target human-triggered workflows and SaaS automations. Smithery, mcp.so, and Glama are directories that index third-party MCP servers. Pipeworx is a single live MCP gateway running 419 data sources behind one protocol, designed for autonomous-agent traffic and on-demand tool loading rather than human-in-the-loop tasks.
Why does on-demand tool loading matter for AI agents?
Loading every available tool definition into model context burns tokens on definitions the agent never uses — what we call the context tax. Pipeworx returns only the ~20 most relevant tools per task via discover_tools, ?task= scoping on the gateway URL, or vertical bundles. Compound _intel tools further collapse 5–15 round-trips into one. Result: less context burned, faster responses, more stable transcripts.
Is Pipeworx free to use?
Yes for most usage. The free tier allows 50 calls per day anonymously or 2,000 calls per day with a free GitHub-linked account. Paid plans add unlimited usage with metered billing, budget controls, and access to premium verticals (Housing Intel, FinTech, PharmaPulse, GovCon Radar, AgriPulse, TradeGuard, GreenLens). All public-source data is included on every tier.
Get started
Use the SDK in any TypeScript/Node project, or connect any MCP client — Claude Desktop, Cursor, Windsurf, or your own app.
import { Pipeworx } from '@pipeworx/sdk';
const px = new Pipeworx();
const result = await px.ask("What is the US trade deficit with China?");{ "mcpServers": { "pipeworx": { "url": "https://gateway.pipeworx.io/mcp" } } }