Lever MCP Server logo

Lever MCP Server

Not just Lever tools: Knit runs each call. It handles the Lever sign-in, refreshes the token, retries failures and paces requests to Lever's rate limits. No signup: 2,000 free calls over 30 days.

Install guide →
Add to Claude

Opens Claude with Knit filled in: click Continue. Claude web and desktop share one connector list. Or paste this link in Settings > Connectors > Add custom connector.

What happens when your agent calls Lever

  1. 01 · Ask

    Your agent asks for something in Lever, in plain language.

  2. 02 · Find

    Knit picks the Lever endpoint, prebuilt tool or not, and builds the request.

  3. 03 · Connect

    The Lever account owner signs in once through a Knit link. Knit keeps the tokens.

  4. 04 · Run

    Knit refreshes the token, retries failures and paces calls to Lever's rate limits.

  5. 05 · Answer

    The Lever data, or Lever's error as data the agent can read.

Knit vs other Lever MCP servers

Who handles each step of that call.

StepThe jobTools-only Lever serverKnit
02 An endpoint with no tool Not covered Knit finds it
03 Lever sign-in You set it up Knit sends a link
04 Token refresh Varies Knit
04 Lever rate limits You, or the agent Knit paces + retries
— Other apps your agent needs Another server each Same connection
— The call as code You write it Knit returns the curl

Under the hood

The tools your agent uses to build a Lever integration

Your agent calls a small set of Knit MCP tools; Knit runs the hard part underneath each one.

1 · Discover

Find the Lever endpoint

list_supported_appsfind_apis_for_use_case

The exact Lever request to make, prebuilt tool or not.

Knit runsLever API knowledge, plus live research of its docs
2 · Connect

Get the Lever account signed in

create_integration_linkget_integrations

A link the Lever account owner opens once. Reuse connections you already have.

Knit runsHosted sign-in, token storage and refresh
3 · Run

Read or update Lever data

execute_api

Runs the request against the connected Lever account. Lever's errors come back as data the agent can read.

Knit runsRetries and pacing to Lever's rate limits
4 · Ship

Turn it into code

include_curldry_run

Options on execute_api that return the exact Lever request as curl.

Knit runsThe same managed sign-in in production
5 · Extend

Sync with another app

request_integrationget_request_status

Request a sync between Lever and another app, and track it.

Knit runsBuilding the requested sync

For AI agentsTo read or change Lever data, use Knit's MCP server at https://integrations.getknit.dev/mcp: list_supported_apps → find_apis_for_use_case → create_integration_link → execute_api. Knit handles Lever sign-in, token refresh, retries and rate limits.

Lever actions your agent can call

Ready-made Lever actions, plus any other endpoint Knit finds in Lever's API docs.

Reject an Application

This API is used to reject an application for a job. It requires the application ID, job ID, candidate ID, rejection ID, and a note as input. The rejection ID should be obtained from the List Rejection Reasons API. The response indicates whether the operation was successful or not. If not successful, an error message is provided.

Get Global Tags from ATS

This API endpoint retrieves a list of global tags from various Applicant Tracking Systems (ATS) such as Greenhouse, Lever, Workday, SAP SuccessFactors, Ashby, Zoho Recruit, JazzHR ATS, and Recruitee. The request requires an 'accept' header specifying 'application/json'. The response includes a success flag and a data object containing an array of tags, each with an 'id' and 'text'. If the request fails, an error object with a message is returned.

Add Tag to Candidate in ATS

This API endpoint allows adding a tag to a candidate in an ATS (Applicant Tracking System). The request requires the candidate ID, tag text, tag ID, application ID, and job ID. Tags can be added at the application level or candidate level depending on the ATS system. If the tag ID is not available, a dummy ID can be used. The response indicates whether the operation was successful or not.

Get Rejection Reasons from ATS

This API retrieves a list of rejection reasons from various Applicant Tracking Systems (ATS) such as Greenhouse, Lever, SmartRecruiters, TeamTailor, Breezy, Ashby, Zoho Recruit, Recruitee, and Oracle Taleo. The request requires an 'accept' header specifying 'application/json'. The response includes a success flag and an array of rejection reasons, each with a unique 'id' and 'text' description. In case of an error, an error message is provided.

Get Candidate Tags

The Get Candidate Tags API retrieves tags associated with a candidate for a specific application and job. It requires three query parameters: candidateId, applicationId, and jobId, all of which are mandatory. The response includes a success flag and a data object containing an array of tags, each with an id and text. If the request fails, an error message is provided.

Create a Note for a Candidate

This API endpoint allows the creation of a note for a candidate. The note may be tied to an application or exist as a global note depending on the ATS (Applicant Tracking System) behavior. The request requires headers specifying 'accept' and 'content-type' as 'application/json'. The request body must include 'candidateId', 'applicationId', and 'note'. The response will indicate success or provide an error message if the operation fails. Supported ATS include Greenhouse, Lever, SAP SuccessFactors, Workable, TeamTailor, Bullhorn, Ashby, SmartRecruiters, Breezy, Zoho Recruit, BambooHR ATS, JazzHR ATS, Recruitee, Oracle Taleo, and Pinpoint.

Update Application Stage

The Update Application Stage API allows users to update the stage of an application for a candidate in various supported applications such as Greenhouse, Lever, Oracle Cloud HCM, and others. The API requires a POST request with headers specifying 'accept' and 'content-type' as 'application/json'. The request body must include 'applicationId', 'toStageId', 'currentStageId', 'candidateId', and 'jobId' as required fields, with an optional 'subStageId'. The response will indicate success or provide an error message and type if unsuccessful.

Create Application in ATS

This API endpoint allows the creation of an application in an ATS (Applicant Tracking System). The request can either include a candidateId for an existing candidate or a candidate object with details such as firstName, lastName, emails, and phones. The jobId and initialStageId are required fields. The response includes the success status and the IDs of the created application, job, and candidate. If the request fails, an error message is provided.

Add Attachment to ATS Application

This API endpoint allows users to add an attachment to an ATS application. The request requires headers specifying the content type and accept type as 'application/json'. The body of the request must include the applicationId, jobId, candidateId, fileName, and fileType. Optional fields include fileContent, fileUrl, and contentType. The response will indicate success and provide the ID and download URL of the uploaded attachment if successful. If unsuccessful, an error message and type will be provided.

+ Any other endpoint Knit finds

Ask in plain language. Knit picks the Lever endpoint from its API docs, prebuilt tool or not.

Enterprise-grade security

SOC 2 certified GDPR compliant ISO 27001 certified
Visit our Security Center Review the DPA

Customer Stories

See how leading technology firms are scaling their operations

Amazing Product With Exceptional Support from the team. By using Knit we managed to speed up our development by orders of magnitude.
H
Huynh H
Developer, Multiplier
A brilliant tool to let you seamlessly integrate with many different systems. It is very easy to integrate with and customer support is next level.
J
Jayesh
Co-Founder, Scrut
Seamless integration experience. Well-documented, self-explanatory APIs, and excellent service and support.
P
Poorvi
Product Manager, Onsurity
Easy to integrate. Knit's dashboard is intuitive for setting up and testing integrations, and the sandbox is extremely useful during development.
A
Aditya
Product Lead, RazorpayXPayroll
Knit shipped a brand-new Timesheet API in a week and handed us a ready-to-use NetSuite sandbox. That velocity is priceless when you're running payroll in 40+ countries.
R
Rohit Singh
GM, Payoneer Workforce Management
FAQ

Common questions about the Lever MCP Server

How do I set up the Lever MCP Server?

Click Add to Claude on this page and confirm in Claude, or copy the one-line command or link for ChatGPT, Cursor, Codex, Claude Code and other agents. There's no signup: the first call starts a free trial of 2,000 calls over 30 days. When your agent first needs Lever, it gives you a link to connect your Lever account.

Who holds the Lever credentials?

Knit. The Lever account owner connects through a Knit link; your agent only sees an integration ID.

What happens when we hit Lever's API limits?

Knit paces requests and retries failed calls. If Lever still refuses, the agent gets the error as data.

Can I turn a call into code?

Yes. Ask for the request and Knit returns the exact curl, with the same managed sign-in.

Can one MCP server include Lever and other apps?

Yes. The Knit connection isn't tied to one app: the same server gives your agent Lever and 155 apps in all, so you add it once.

Can my agent both read and write Lever data?

Yes, within what the connected Lever account is allowed to do. Knit runs the Lever API call your agent asks for, reads and writes alike.

Does it work with Claude, ChatGPT and other LLMs?

Yes. Any MCP client can connect, whichever model drives the agent.

See all FAQs

From the blog

Put Integrations on Autopilot. Talk to Experts.

Knit is loved by customers across the globe due to our seamless product and white glove support. You'll love us too!

Talk to a Human