Source URL
Where it came from
The source URL and source name travel with the Manifest so an agent can cite or inspect the original source event.
Provenance
Source-Grounded Content API.
A source-grounded content API gives an AI system structured content plus the source evidence behind it. The point is not just to retrieve text; it is to keep source identity, URLs, dates, lineage, and stable IDs attached to every usable item.
Synorb returns Manifests from watched Source Channels and Streams. Each Manifest can carry Briefs, Signals, Records, source URLs, provenance, and delivery metadata for REST, MCP, webhooks, and S3.
Source URLs · Stable IDs · Stream routing · Briefs / Signals / Records
01 / Grounding
Grounding is more than a link.
For AI applications, a citation is useful only if the system can keep it attached through retrieval, summarization, review, and workflow routing. Synorb treats provenance as part of the object, not as an afterthought.
Dates
When it happened
Published dates and capture dates help the application reason about freshness, ordering, and source timing.
Lineage
How it entered
Source Channel, Stream routing, Record lineage, and stable IDs keep downstream joins and audits possible.
02 / Object
The API unit is a Manifest.
Manifests let an agent use content without separating the claim from its evidence. A single Manifest can support citation display, retrieval indexing, alert generation, human review, and warehouse joins.
Manifest excerptprovenance
{
"manifest_id": "1777525429698648000",
"source": {
"name": "Observed Source Channel",
"url": "https://source.example/update",
"published_date": "2026-06-17",
"captured_at": "2026-06-17T14:05:00Z"
},
"brief": {
"title": "Source event summary"
},
"signals": [
{
"claim": "Atomic claim grounded in the source event",
"evidence": "Text span or source-grounded evidence reference"
}
],
"record": {
"stable_id": "record_1777525429649909800"
},
"lineage": {
"source_channel_id": "observed-source-channel"
}
}
03 / Workflows
Use source-grounded content where trust matters.
Source-grounded content is useful when a model output needs to survive review: alerts, research dashboards, company monitoring, current-event feeds, newsletters, RAG pipelines, and agent actions that cite evidence.
AI apps
Cited user experiences
Show source URLs next to summaries, rankings, answers, and feed cards so users can inspect the evidence.
Data teams
Repeatable joins
Use stable IDs, Records, and Stream routing to move the same source event through databases, warehouses, and evaluation jobs.
04 / Connect
MCP for agents, REST for systems.
The same source-grounded data can be used interactively and in production. MCP lets agents query Streams and retrieve Manifests. REST, webhooks, and S3 fit backend-owned applications, event pipelines, and archival workloads.
05 / Limits
Grounded does not mean unlimited coverage.
Synorb is not arbitrary open-web scraping. It is a source-grounded feed system for watched Source Channels and configured Streams. For unknown sources or broad exploration, use web search alongside Synorb.
| Synorb is good for | Use search for |
|---|---|
| Known source coverage, repeated monitoring, cited AI outputs, stable IDs, data pipelines, and source-aware agent memory. | Finding unknown sources, ad hoc browsing, broad background research, and questions outside watched Streams. |
06 / Pricing
Use the canonical pricing page.
Streams pricing explains the current plans, quotas, overage rules, and delivery cadence. Link there when evaluating cost instead of relying on copied pricing text.
FAQ
The short version.
What is a source-grounded content API?
A source-grounded content API returns structured updates with the underlying source identity, source URL, dates, lineage, and provenance attached so an AI system can cite and audit what it used.
What provenance does Synorb include?
Synorb Manifests can include source URL, source name, published date, capture date, Source Channel, Stream routing, Record lineage, stable IDs, and ontology tags.
Can an agent cite Synorb results?
Yes. Synorb keeps source URLs and source metadata attached to Manifests so agents and applications can present citations or trace an output back to the original source event.
Is Synorb a RAG database?
No. Synorb supplies source-grounded content feeds and Manifests that can feed retrieval systems, agent memory, warehouses, alerts, and review tools. It is not a replacement for your application database.
What are the limits?
Synorb is strongest for watched Source Channels and configured Streams. Use web search for unknown sources, open-ended discovery, or questions outside watched coverage.
Start
Give the model something it can trace.
Start with keys, then connect through REST, MCP, webhooks, or S3 depending on whether the work is interactive or production-owned.