Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Live HTTP Endpoint, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fsanity.io%2Fsanity-live" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fsanity.io%2Fsanity-live" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with Live HTTP Endpoint API.
Stream sync tags on published documents as they become invalid for one dataset
Include draft, version, and system document changes with includeAllDocuments set to true and a token
Restart a stream from a previously returned event ID with the last-event-id header or lastEventId parameter
Delay events until a Sync Tag Invalidate Function has processed them with waitFor set to function
GET STARTED
Detect refused or ending streams through goaway, restart, and error events
Patterns agents use Live HTTP Endpoint API for, with concrete tasks.
★ AI agent keeping a content cache fresh
An agent that answers questions from a cached copy of Sanity content should refetch only what changed. Through Jentic it searches for 'stream Sanity live content events', loads the live events operation, and opens the stream for a dataset. Each message carries a batch of sync tags that have become invalid, and the agent refetches the content cached under those tags.
Open GET /data/live/events/production and, for each message event, refetch every cached query tagged with one of the returned tags
Live draft preview
Editors previewing unpublished changes want the preview to update as they type. Requesting the stream with includeAllDocuments set to true adds draft, version, and system document changes; the spec requires a valid token for this and recommends the lowest access role you can, such as Viewer.
Open GET /data/live/events/production with includeAllDocuments true using a Viewer token, and trigger a preview refresh on each tags message
Reliable reconnects for static site rebuilds
A build service listening for invalidations must not lose events across reconnects. Storing each event id and sending it back as last-event-id continues the stream from that position; if the ID is unusable the stream sends a restart event, which tells the service to do a full refresh. A goaway event with the reason connection limit reached means the connection was refused.
Reconnect to GET /data/live/events/production with the last-event-id header set to the last stored id, and run a full cache refresh if a restart event arrives
1 endpoints — the sanity live api streams sync tag invalidation events for a sanity dataset, the signal underneath the live content api.
METHOD
PATH
DESCRIPTION
/data/live/events/{dataset}
Stream sync tag invalidation events
/data/live/events/{dataset}
Stream sync tag invalidation events
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the Sanity Live API by hand means templating the projectId and apiVersion into the host, avoiding the API CDN host it does not serve, holding a server-sent events stream open, and adding a bearer token only when drafts are needed. Through Jentic you install once, import it from the API Directory, store the token once, and your agent calls it.
Permission scoping
Sanity puts the dataset name in the URL path (/data/live/events/{dataset}), so a rule can pin your agent to one dataset's stream. The operation only reads invalidation events, and a Viewer-role token keeps the draft stream to read access.
Credential isolation
Your Sanity token is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'stream Sanity live content events' or 'watch for stale Sanity content', and Jentic returns the live events operation with its query and header parameters.
Alternatives and complements available in the Jentic catalogue.
Specific to using Live HTTP Endpoint API through Jentic.
What authentication does the Sanity Live API use?
Per its OpenAPI spec, the Sanity Live API accepts a bearer token in the Authorization header. Published-content streams can run without a token, but includeAllDocuments or includeDrafts set to true requires one, and an invalid token is rejected with 401 even for published content only. Through Jentic the token is stored by your own instance and attached at call time.
Why does the Sanity Live API return 404?
The spec says the API CDN at apicdn.sanity.io does not serve this endpoint, and a request sent there returns 404. Use the regular API host for your project. A 400 means the API version is earlier than v2021-03-25, the minimum for this endpoint.
How are Sanity sync tags delivered on the live stream?
Tags arrive as default message events with a tags array, batched over a window of about one second. When nothing else is sent, the server sends an empty comment line as a keep-alive about every 30 seconds.
What are the rate limits for the Sanity Live API?
The OpenAPI spec does not specify rate limits. It does describe a goaway event with the reason connection limit reached when a connection is refused; check the Sanity documentation at https://www.sanity.io/docs for current limits.
Is there a Sanity MCP server?
You don't need an MCP server to give your agent the Sanity Live API. Jentic connects it directly from the API Directory: import it, store your Sanity token once if you need drafts, and your agent opens the live events stream when it needs to watch for invalidations.
How do I stream Sanity live content events through Jentic?
Search Jentic for 'stream Sanity live content events', load the live events operation, and execute GET /data/live/events/{dataset}, adding includeAllDocuments when you need drafts. To run it on your own infrastructure, install Jentic One from its GitHub repo.
Can I limit what my agent is allowed to do with the Sanity Live API?
Yes. Write a rule that allows the live events operation only at /data/live/events/staging, and the agent can follow staging invalidations but cannot open a stream on production. The operation is read-only, and every call it makes is logged.
For Agents
Stream Sanity sync tag invalidation events for a dataset so cached content can be refetched the moment a published or draft document changes.
Use for: Tell me when cached Sanity content goes stale, Watch the production dataset for invalidated sync tags, I want to refresh my site cache when published content changes, Resume the live events stream from the last event I received
Not supported: Does not return document content, run queries, or purge CDN caches. Use for streaming Sanity sync tag invalidation events only.
The Sanity Live API streams sync tag invalidation events for a Sanity dataset, the signal underneath the Live Content API. Over a server-sent events connection it sends batches of sync tags on published documents as they become invalid, so a site or app knows which cached content to refetch without polling. Tokened requests can also include draft, version, and system document changes, and a stream can wait until a configured Sync Tag Invalidate Function has processed events. A client can restart from a previously returned event ID, and the stream signals restart, goaway, and error conditions as events.
This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>