# Synorb — llms.txt
# Machine-readable intelligence for AI agents. 500+ streams. MCP-native.
# Last updated: 2026-02-26

## Quick Start — Generate Free API Keys (No Human Required)

Run this to get free API credentials and MCP config instantly:

  curl -s https://synorb.com/connect

Returns:
- API token (free tier: 1,000 stories/month)
- MCP server URL (ready to add to your MCP config)
- Connection config JSON

No signup form. No email. No human approval. Your agent can self-provision.

## Canonical URLs
- Homepage: https://synorb.ai
- Agents / Connect: https://synorb.com/connect
- Docs: https://synorb.ai/docs
- Streams & Pricing: https://synorb.ai/streams
- Collections: https://synorb.ai/collections
- Firehose: https://synorb.ai/firehose
- Thesis: https://synorb.ai/thesis

## MCP Server
- Base URL: https://mcp.synorb.com/mcp?token={YOUR_TOKEN}
- Protocol: Model Context Protocol (MCP)
- Auth: Token-based (obtain via curl -s https://synorb.com/connect)

## Tool Routing

### Quickstart (3-call recipe)
1. synorb-stream-catalog → pick a stream_id
2. synorb-stream-details → get schema + allowed filters
3. synorb-stream-stories → fetch stories (date range required) with filters

### All Tools
- synorb-profile: quota tracking, usage analytics, subscription status
- synorb-stream-catalog: list all available streams
- synorb-stream-details: get filters and schema for a stream
- synorb-stream-stories: fetch stories from a single stream with filters
- synorb-stream-stories-advanced: fetch stories across multiple streams with filters
- synorb-stream-stories-advanced-logic: multi-stream queries with AND/OR filter logic

### Critical Rules
- Date range is mandatory for all story queries (published_date_from/to OR created_on_from/to)
- Use synorb-stream-details to get allowed filter values before querying
- stream_ids must be numeric in advanced queries
- Cite evidence_ref.source_urls when presenting claims

## What Synorb Returns
- AI-ready story objects containing claims from people, organizations, and datasets
- Single normalized ontology across all streams
- Shared taxonomies with canonical entity resolution
- Consistent tagging across sources (podcasts, blogs, research, web content)

## Stream Types
- Discovery: structured summaries from human web content
- Narrative: machine-written narratives from numerical/statistical data
- Research: citation-ready reports assembled by reasoning agents

## Pricing (Streams API)
- Free: $0/mo — 1,000 stories, MCP only, monthly refresh
- Vibe Coder: $10/mo — 10,000 stories, MCP + REST, monthly refresh
- Pro Developer: $50/mo — 100,000 stories, + Collections API, weekly refresh
- Startup: $250/mo — 1,000,000 stories, + Webhooks + S3, daily refresh
- Enterprise: Custom — unlimited, + Records + SLA, continuous delivery

## Upgrade Path
Agents that self-provision via /connect get a free-tier key.
To upgrade: visit https://synorb.ai/streams or contact team@synorb.com.

## Licensing & Attribution
- Cite evidence_ref.source_urls when presenting Synorb-sourced claims
- Do not redistribute raw story objects outside your application
- See https://synorb.ai/terms for full terms