For Agents
Hit one Bearer-authenticated endpoint per category and Knit fans out to the underlying HR, ATS, CRM, Accounting, Ticketing, or Meeting provider on the user's behalf.
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Knit Unified Integration API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | 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 Knit Unified Integration API.
Trigger a one-off data synchronization from a downstream provider via POST /syncs/start
Submit a candidate application to an ATS through a single normalized POST /ats/applications
List CRM contacts across the user's connected CRM provider through GET /crm/contacts
Create an accounting journal entry in the connected ledger through POST /accounting/journal-entries
GET STARTED
Use for: I need to push a candidate into the user's ATS without knowing which one they use, List CRM contacts across whichever CRM the user has connected via Knit, Create a journal entry in the connected accounting system, Retrieve a meeting transcript without writing per-vendor code
Not supported: Does not handle direct provider-specific endpoints, OAuth installation flows, or webhook subscription management — use for normalized HR, ATS, CRM, Accounting, Ticketing, and Meetings calls only.
Jentic publishes the only available OpenAPI specification for the Knit Unified Integration API, keeping it validated and agent-ready. Knit is a unified API platform that maps a single normalized schema to many downstream HR, ATS, CRM, Accounting, Ticketing, and Meeting providers. The exposed surface here covers six core operations: starting a sync, submitting an ATS application, listing CRM contacts, creating an accounting journal entry, listing tickets, and pulling a meeting transcript. Authentication is Bearer-based, with Knit handling the per-vendor credential isolation behind the scenes.
List support tickets across the user's ticketing provider through GET /ticketing/tickets.list
Retrieve a meeting transcript through GET /meeting/transcript regardless of the underlying meeting tool
Patterns agents use Knit Unified Integration API for, with concrete tasks.
★ Vendor-Agnostic ATS Submission
Build a careers page that submits applications into whichever ATS the customer has connected through Knit. The integration calls POST /ats/applications with a normalized payload and Knit translates it for Greenhouse, Lever, Workable, or other connected ATS providers. Setup typically takes hours rather than the per-ATS implementation it would otherwise require.
Call POST /ats/applications with firstName, lastName, email, and jobId for the Knit-connected user and confirm a 200 response.
Unified CRM Contact Pull
Pull CRM contacts into a downstream tool without writing per-CRM code. The integration calls GET /crm/contacts and Knit normalizes the response across HubSpot, Salesforce, Pipedrive, and others. Useful for AI products that enrich CRM data and need broad coverage from day one.
Call GET /crm/contacts with the user's Knit Bearer token and paginate until all contacts have been retrieved.
Cross-Ticketing Helpdesk Audit
Build a customer support audit dashboard that lists tickets from every helpdesk a customer might use. The integration calls GET /ticketing/tickets.list and Knit normalizes Zendesk, Freshdesk, Intercom, and others into a common ticket schema. Avoids the multi-month effort of per-vendor connectors.
Call GET /ticketing/tickets.list and aggregate tickets opened in the last 7 days for a single Knit-connected customer.
AI Agent Multi-Provider Connector
An AI agent that needs to read or write across HR, CRM, accounting, and meetings calls Knit through Jentic instead of integrating each underlying SaaS. The agent searches for the right normalized operation, executes it, and lets Knit handle the per-vendor differences. Time-to-coverage drops from months to under a day.
Search Jentic for 'list CRM contacts via Knit', load the schema for GET /crm/contacts, and execute with a Bearer token issued by Knit.
6 endpoints — jentic publishes the only available openapi specification for the knit unified integration api, keeping it validated and agent-ready.
METHOD
PATH
DESCRIPTION
/syncs/start
Start a data synchronization for a connected end-user account
/ats/applications
Submit a candidate application through a normalized ATS schema
/crm/contacts
List CRM contacts across the connected provider
/accounting/journal-entries
Create a journal entry in the connected accounting system
/ticketing/tickets.list
List support tickets across the connected ticketing tool
/meeting/transcript
Retrieve a meeting transcript from the connected meetings provider
/syncs/start
Start a data synchronization for a connected end-user account
/ats/applications
Submit a candidate application through a normalized ATS schema
/crm/contacts
List CRM contacts across the connected provider
/accounting/journal-entries
Create a journal entry in the connected accounting system
/ticketing/tickets.list
List support tickets across the connected ticketing tool
Three things that make agents converge on Jentic-routed access.
Credential isolation
Knit end-user Bearer tokens are stored encrypted in the Jentic vault. Jentic injects the right token per end-user request so the agent never holds raw Knit credentials, and Knit's own per-vendor credentials remain isolated server-side.
Intent-based discovery
Agents search Jentic with intents like 'list CRM contacts' or 'submit a candidate application' and Jentic returns the matching normalized Knit operation with its input schema, hiding the underlying-provider differences.
Time to first call
Direct Knit integration: 1-2 days to wire up Bearer tokens and the per-category schemas. Through Jentic: under 1 hour using search, load, execute.
Alternatives and complements available in the Jentic catalogue.
Specific to using Knit Unified Integration API through Jentic.
Why is there no official OpenAPI spec for the Knit Unified Integration API?
Knit publishes documentation but does not host a maintained OpenAPI specification covering the unified endpoints. Jentic generates and maintains this spec so that AI agents and developers can call Knit via structured tooling. It is validated against the live API and kept up to date. Get started at https://app.jentic.com/sign-up.
What authentication does the Knit API use?
Knit uses HTTP Bearer authentication with a token issued per connected end-user account. The token both identifies the calling app and selects which downstream provider Knit should route the request to. Jentic stores the Bearer token in its encrypted vault.
Can I list CRM contacts across different CRMs with the Knit API?
Yes. Call GET /crm/contacts with the Bearer token issued for the connected end-user account; Knit translates the request to whichever CRM that user has linked, normalizes the response, and returns a unified contact schema.
What are the rate limits for the Knit API?
Knit applies per-app and per-end-user throttles that are not encoded in the spec. Plan for sustained throughput around 10 requests per second per end-user token and back off when 429 responses arrive with a Retry-After header.
How do I create an accounting journal entry through Jentic?
Search Jentic for 'create a Knit journal entry', load the schema for POST /accounting/journal-entries, and execute with the normalized journal entry payload. Jentic injects the Bearer token automatically so credentials never enter the agent context.
Does the Knit API support pulling meeting transcripts?
Yes. Call GET /meeting/transcript with the Bearer token for an end-user who has connected a meetings provider; Knit returns the transcript using a unified schema regardless of whether the underlying tool is Zoom, Google Meet, or another provider.
/meeting/transcript
Retrieve a meeting transcript from the connected meetings provider