Is Synorb a web scraping API?

No. Synorb is built around watched Source Channels, Streams, and Manifests. It helps agents listen to maintained coverage areas instead of scraping arbitrary pages at query time.

What does Synorb return?

Synorb returns Manifests that can include Briefs, Signals, Records, source metadata, source URLs, published dates, stable IDs, tags, and usage information.

Can a shipped app use this API?

Yes. Shipped apps should call Synorb REST from server-side code, accept webhooks, or ingest S3 exports. MCP is best for agent-native development and interactive workflows.

Developer intent

Structured content API for AI agents.

Q: What is a structured content API for AI agents?

It is an API that returns machine-readable content objects with source links, provenance, stable IDs, timestamps, and fields an agent can store, route, summarize, or cite.

AI apps need fresh content that arrives as objects, not pages the model has to clean up by itself. Synorb returns source-grounded Manifests with Briefs, Signals, stable IDs, source URLs, published dates, Stream routing, and delivery metadata.

Get Keys API docs →

API shape

The response is already structured for agents.

A structured content API should reduce parsing work. Synorb returns fields an application can store and render directly: headline, summary, source URL, date, source channel, tags, significance, usage, and pagination.

type ManifestRow = { id: string; headline: string; summary: string; publishedDate: string; sourceUrl: string; sourceName: string; streamNames: string[]; tags: Array<{ type: string; name: string }>; };

Positioning

Not a scrape-anything endpoint.

Synorb is for monitored coverage. It watches Source Channels, organizes them into Streams, and emits Manifests. That is different from asking an agent to fetch arbitrary URLs, classify them, extract claims, dedupe results, and remember what is stale.

Synorb

Maintained feed

Best forknown sources, recurring topics, production apps, cited context

UnitManifest with Brief, Signal, source metadata, stable IDs

Search / scrape

Ad hoc discovery

Best forunknown sites, long-tail questions, one-off browsing

Unitpages or snippets the agent still has to evaluate

Delivery

Use the right interface for the job.

REST API

Use server-side REST routes for shipped apps, dashboards, websites, schedulers, and backend-owned product paths.

MCP

Use MCP when a compatible agent should discover Streams, inspect sample Manifests, and reason through source-backed context.

Webhooks and S3

Use webhooks for event-driven workflows and S3 for durable archive, batch ingestion, and replay.

FAQ

Short answers.

What is a structured content API for AI agents?

An API that returns machine-readable content objects with source links, provenance, stable IDs, timestamps, and fields an agent can store, route, summarize, or cite.

Does it include citations?

Yes. Source URLs and source metadata are part of the Manifest so agents can cite the originating source, not just a generated summary.

How fresh is the data?

Freshness depends on plan and Stream cadence. Use the current pricing pages for the canonical plan and cadence details.

Start

Build against objects, not scraped pages.

Get Keys Build guides → Search alternative →