BambooHR and AI

MCP Server

BambooHR's AI connectors are currently in a closed beta. Interested in joining? Request access.

The BambooHR MCP server lets you connect any Model Context Protocol-compatible AI tool or agent to your BambooHR account. It's the foundation that powers the BambooHR Claude connector and ChatGPT app, and it's also available directly for developers building custom integrations or connecting MCP clients we don't yet support out of the box.

What it does

The MCP server exposes 44 tools across five core areas of BambooHR: employee data and files, time off and availability, reporting and datasets, hiring and applicant tracking, and goals and performance. We picked this starting set by looking at the most common jobs HR teams want help with. We'll add more tools over time as we learn what customers ask for most.

Once connected, an AI assistant can:

  • Answer questions about employees, time off, goals, and hiring using natural language
  • Pull data from BambooHR datasets and saved reports
  • Take actions on your behalf (create time off requests, update employee fields, manage goals) when you have the right permissions
  • Combine BambooHR data with other tools the same AI client is connected to

Sample prompts

Here are a few ways you can put the server to work. This isn't an exhaustive list, just a starting point:

  • "Who's out of office this week?"
  • "What's Alice's remaining PTO balance?"
  • "Pull last month's headcount report and summarize it"
  • "Show me all open requisitions in the Engineering department"
  • "How many candidates are in interview stage right now?"
  • "Create a time off request for me from June 10 to June 14"
  • "List goals owned by the marketing team and which ones are overdue"
  • "Update the work phone number on file for [employee name]"

Server endpoint

https://{your-company}.bamboohr.com/api/mcp

Replace {your-company} with your BambooHR subdomain. Each customer has their own MCP endpoint scoped to their account.

Authentication

The MCP server supports two authentication methods.

OAuth is the right choice for most use cases. The BambooHR Claude connector and ChatGPT app both use OAuth behind the scenes. You click Connect, sign in to BambooHR, and token management happens automatically.

If you're connecting from a different MCP client today, you'll need to configure the OAuth flow manually. We're working on Dynamic Client Registration (DCR) support so other MCP clients can self-discover and connect without setup. We'll update this page when it's ready.

Bearer token

You can also generate a token manually and pass it as a bearer token. This is useful for testing or scripted access, but keep in mind:

  • Tokens currently have a 1-hour lifetime and need to be refreshed manually
  • Tokens carry the permissions of the user who created them

For long-running or production use, OAuth is the better path.

Prerequisites

Before connecting, make sure you have:

  • An active BambooHR account
  • Permission to access the data you want the AI tool to see (the server enforces BambooHR's existing permissions, so it can't grant access you don't already have)
  • An MCP-compatible client (Claude, Cursor, your own agent, etc.) or the ability to make OAuth-authenticated HTTPS requests

Access model

The MCP server respects BambooHR's permissions exactly. What you can do through an AI assistant matches what you can do when logged into BambooHR directly.

  • HR admins get full company-wide access. This is the most powerful experience and what we've optimized for first.
  • Managers can see and act on data for themselves and their direct reports.
  • Individual contributors can see and act on their own information.

Future updates may expand non-admin access to include basic info about other employees (the same kind of directory info you'd see logged into BambooHR), but for now the IC and manager experience is intentionally scoped.

Guidelines for safe use

A few practices we strongly recommend any time you're connecting BambooHR data to an AI assistant:

  • Use trusted, well-established MCP clients. Stick to clients you know and trust (Claude, ChatGPT, Cursor, etc.). Unverified clients can expose your data in ways you don't intend.
  • Keep your session focused. If you mix the BambooHR MCP server with several other connected servers in the same session, data can move between them in ways that are hard to predict. For sensitive HR work, keep the session scoped to BambooHR.
  • Require confirmation before tool calls run. Most MCP clients let you approve each tool call before it executes. Leave that turned on, especially for any write action (creating time off requests, updating employee data, modifying goals).
  • Verify AI output before acting on it. Treat the assistant's responses as a starting point, not a source of truth. Double-check numbers, names, and policy interpretations before making decisions.
  • Opt out of model training in your AI client. Most providers offer a setting to exclude your conversations from training data. Turn it on. HR data is sensitive and shouldn't be used to train third-party models.
  • Protect bearer tokens. If you're using a manually generated token, treat it like a password. Don't paste it into untrusted tools or commit it to source control.

Available tools

The server organizes tools into five domains. We'll continue adding tools based on customer feedback. If you'd like to see something that isn't here yet, let us know.

Employee data and files

The foundation. Look up employee details, access tabular records like job history or compensation, and see what files are attached to employees or stored at the company level.

ToolDescription
get-employeeGet a single employee by ID
list-employeesList employees in the company
update-employeeUpdate fields on an employee record (write)
get-employee-table-dataGet tabular data (job, compensation, etc.) for an employee
list-employee-filesList files attached to an employee
list-company-filesList company files and their categories
list-fieldsList the fields available on the employee record

Time off and availability

One of the highest-traffic areas in BambooHR. Check who's out, view balances, submit PTO requests, and approve or deny requests as a manager.

ToolDescription
list-whos-outList who's out of office
list-time-off-requestsList time off requests
get-time-off-balanceGet an employee's time off balance
list-time-off-typesList available time off types
list-time-off-policiesList time off policies
list-employee-time-off-policies-v1.1List policies assigned to an employee
create-time-off-requestCreate a new time off request (write)
update-time-off-request-statusApprove or deny a time off request (write)

Reporting and datasets

Where LLM connectors really shine. Build ad-hoc queries using the Datasets API or pull saved reports built in the BambooHR UI.

ToolDescription
list-datasets-v1-2List available datasets
get-fields-from-dataset-v1-2Get the fields available in a dataset
get-field-options-v1-2Get the options available for a field
get-data-from-dataset-v2Run a query against a dataset
list-reportsList saved reports
get-report-by-idGet the results of a saved report
list-tabular-fieldsList fields available in tabular records
list-list-fieldsList fields configured as list types

Hiring and applicant tracking

A natural fit for conversational queries. Hiring managers can check pipeline status, review candidates, and add feedback without switching tools.

ToolDescription
get-job-summariesGet summaries of open jobs
get-applicationsGet job applications
get-application-detailsGet details for a specific application
get-statusesGet applicant statuses available in the pipeline
create-application-commentAdd a comment to an application (write)

Goals and performance

Full lifecycle management for goals. Create, update, close, and reopen goals, track progress at the goal and milestone level, manage comments, and filter or align goals across the organization.

ToolDescription
list-goalsList goals
get-goals-aggregate-v1.2Get aggregate goal data across the company
get-goal-aggregateGet aggregate data for a single goal
get-goals-filters-v1.2Get available goal filters
get-alignable-goal-optionsList goals available for alignment
list-goal-commentsList comments on a goal
create-goalCreate a new goal (write)
update-goal-v1.1Update an existing goal (write)
update-goal-progressUpdate progress on a goal (write)
update-goal-milestone-progressUpdate progress on a goal milestone (write)
create-goal-commentAdd a comment to a goal (write)
update-goal-commentUpdate a goal comment (write)
delete-goal-commentDelete a goal comment (write)
close-goalClose a goal (write)
reopen-goalReopen a previously closed goal (write)
delete-goalDelete a goal (write)

How it works

When an MCP client calls one of the server's tools, here's what happens:

  1. The client sends an MCP tool call to https://{your-company}.bamboohr.com/api/mcp with an OAuth access token or bearer token
  2. The server authenticates the request and identifies the user
  3. It validates the request and checks that the user has permission to perform the action
  4. It calls the appropriate BambooHR API endpoint(s) on the user's behalf
  5. It returns the result back to the client in MCP format

The server is hosted by BambooHR. There's nothing to install or run locally.

Limitations

  • The server exposes about 40 endpoints today. Not every BambooHR API capability is available yet.
  • Bearer tokens require manual refresh every hour.
  • DCR isn't yet supported for self-discovery from arbitrary MCP clients. OAuth setup may require manual configuration in clients that don't ship with a pre-built BambooHR connector.
  • The MCP server can read and write within the scope of the connected user's permissions, but some sensitive actions (like changes that require approvals or workflows inside BambooHR) still go through BambooHR's existing flows.

What's next

We're actively expanding the server. Upcoming work includes:

  • More tools across additional BambooHR areas
  • DCR support for easier setup from any MCP client
  • Longer token lifetimes and refresh handling improvements

If you have feedback or run into something that doesn't work the way you expect, reach out to your BambooHR contact or open a support request.

Disclaimer

The BambooHR MCP server connects you to third-party AI tools, which can produce inaccurate or incomplete responses. BambooHR isn't responsible for errors in AI-generated output or for actions taken in your account based on those responses. Always verify information and review actions before confirming them.

On this page