Skip to content

Recruitee MCP server for hiring managers

A Recruitee MCP server for hiring managers needs five read tools. The API token follows its creator's role, so pick the creator first, then narrow it.

Raajshekhar Rajan 4 min read
A hiring manager at a laptop reading a list of interview results and candidate notes that an AI assistant pulled up before a debrief meeting

Can hiring managers use Recruitee in Claude or ChatGPT?

Yes. A Recruitee MCP server for hiring managers means connecting Recruitee in Elaichi with a personal API token, adding https://api.elaichi.ai/mcp in Claude, ChatGPT, Cursor or Microsoft Copilot, and holding hiring managers to the three read tools they need with a restriction on their role.

A hiring manager opens the applicant tracking system before a debrief, reads what interviewers wrote, and closes it. How did the interviews go, what did the team note, which office is the role in? MCP (Model Context Protocol) is the standard way an AI assistant calls tools in other apps, and Elaichi's Recruitee connector offers five tools, small enough to review by hand.

Which Recruitee tools should hiring managers reach?

All five Recruitee tools read, and two of them deserve a second look before a hiring manager gets access. The table shows the split.

Tool What it returns Hiring manager
list_all_recruitee_interview_results Interview results with candidate, admin and evaluation details Yes
list_all_recruitee_notes Notes on one candidate, with author, body and visibility Yes
list_all_recruitee_locations Company locations Yes
list_all_recruitee_offer_notes Notes on an offer, with replies Usually no
list_all_recruitee_admins Admin names, emails and roles No

Offer notes tend to hold pay talk between recruiting and finance. The admin list is a directory of who holds power in the account. Neither helps a manager prepare for an interview.

Every tool needs your Recruitee company ID as an input, so the AI client asks for it or takes it from the instructions you give.

What does a Recruitee API token let the AI see?

It lets the AI see what the token's creator could see. Recruitee says the permissions of a personal API token are based on the Hiring role of the person who created it, and that the token does not expire unless the user revokes it (Recruitee support FAQ, checked November 2026).

Two consequences follow. The token outlives the project that created it, so someone must own revoking it. And whoever makes it decides the ceiling, so a token made by an account admin carries admin reach even if only hiring managers ever use it.

Recruitee sends the token as a Bearer value together with the company ID (Recruitee API quickstart, checked November 2026). Elaichi holds the token and sends it for you, so hiring managers never paste a token into a chat.

Should hiring managers share one Recruitee connection?

Share one connection when every manager should see the same thing. Give each manager their own when Recruitee's Hiring roles already differ by manager. Both are valid, and they behave differently.

Every call through a connection runs on the account the connection was made with. A shared connection acts as its owner for every manager. If an admin created it, the managers get an admin's view, and only Elaichi restrictions narrow it.

Per-manager connections keep the Hiring role of each person. Recruitee then does the narrowing, and Elaichi adds a second layer on top. That is more setup, and it is the safer default for notes, because note visibility is Recruitee's rule to enforce.

Should the manager role use block rules or an allow rule?

Block rules, because the Recruitee connector is five read tools and managers need three. Holding a role to an app's reads takes either an allow rule naming those reads, plus allow rules for every other app the role uses, or block rules naming every tool that is not a read. The test is which list is shorter. Here two tools are withheld, so write one block rule on the hiring manager role naming list_all_recruitee_admins and list_all_recruitee_offer_notes.

An allow rule has a cost: once a role carries any allow rule, every connector and tool none of its allow rules names is blocked for that role. A manager role with one Recruitee allow rule would lose every other app until you allow them too. On the block side, a layer with only block rules has no allowlist, so everything it does not block passes it.

A restriction targets a role or one member, and custom roles and restrictions both come with the Gold and Black plans. A restricted tool is left out of the tools Elaichi offers a client and cannot be called. Search still names it, flagged restricted, with no schema.

Changes take effect within about two minutes.

What does the audit log show after a debrief?

Each call that reaches execution writes one entry for each connected-tool call that reaches execution (a call refused earlier writes none). The row records the person, the tool, the Recruitee connection and the MCP client that made the call.

Because a call from an MCP client is recorded under the person who signed in, with surface mcp and the OAuth client named, a debrief question asked in Claude shows under the hiring manager's name. For the notes and interview tools the row keeps the one path argument that names the object, as the target id, and nothing else about the arguments. That is enough to answer a question such as who opened the notes on one candidate last week.

If a compliance reviewer needs to read this, give them the Auditor role. It is free, read-only, and does not take a billable seat.

How do you set up a Recruitee MCP server for hiring managers?

Each manager adds https://api.elaichi.ai/mcp to their AI client and signs in through the browser. The address is the same for every person. Claude, ChatGPT, Cursor and Microsoft Copilot all connect to it, as does any MCP client that does OAuth in a browser.

On the consent screen they tick Use your connected apps. Then ask: "Show interview results for the product designer role." If a result comes back, the chain works. Then ask for the admin list. The answer should say the tool is restricted for the role, and point at an access request.

Do hiring managers need an AI layer on Recruitee at all?

If hiring managers read Recruitee a few times a month and nobody wants AI help, the Recruitee screens are enough. They already show each Hiring role its own view.

The AI layer helps when a manager wants interview notes beside a calendar and a message draft, in one chat. See the Recruitee connector for its tools, and the guide to allowlists versus blocklists for MCP tools to choose a rule shape.

FAQ

Frequently asked questions

How many tools does the Recruitee connector have?

Five. Elaichi's Recruitee connector offers list_all_recruitee_interview_results, list_all_recruitee_notes, list_all_recruitee_locations, list_all_recruitee_offer_notes and list_all_recruitee_admins. Each one reads data and none creates, edits or deletes anything, so a hiring manager's AI client cannot move a candidate or change an offer through it.

Whose permissions does a Recruitee API token carry?

The permissions of the person who created it. Recruitee says a personal API token is based on the Hiring role of its creator, and that it does not expire unless the user revokes it ([Recruitee support FAQ](https://support.recruitee.com/en/articles/8213076-faq-api), checked November 2026). Treat the token like a standing login for that role and revoke it when the person changes jobs.

Can hiring managers see candidate notes written by other people?

They can see whatever the token's creator can see in Recruitee. The notes tool returns each note with its author, body and a visibility field. Elaichi restrictions name tools, not individual notes, so a role can reach the whole notes tool or none of it.

Is Recruitee enough on its own for hiring managers?

Yes, when managers only read Recruitee and nobody needs a record of AI activity. Recruitee already controls what each Hiring role sees. Add Elaichi when managers also want Gmail, Slack or a calendar in the same chat, or when you need one audit log across apps.

Put AI to work in your own apps

14 days of Black, no credit card. Start with one app and one team.

When the trial ends, nothing is deleted. You move to the free Personal plan, and connections, roles and the audit log stay where they are.

Add one link to your AI app

The same link for everyone on your team.

https://api.elaichi.ai/mcp
  • ChatGPT
  • Claude
  • Cursor
  • Microsoft Copilot