The Loxo MCP (Model Context Protocol) Server allows compatible AI assistants to securely access your Loxo data. Once connected, your AI can search candidates, retrieve jobs, create notes and perform other actions using natural language - without manually copying information between applications.
Important: The Loxo MCP Server supports both read and write actions. Your AI assistant can create, update or permanently delete records, including people, companies and jobs.
We recommend reviewing AI-generated actions before approving them and only connecting trusted AI assistants to your Loxo account.
Setup
Step 1: Copy Your MCP Server URL
Your MCP Server URL follows this format:
https://app.loxo.co/api/{agency_slug}/mcp
Replace {agency_slug} with your Loxo agency slug.
Step 2: Add the Server to Your AI Assistant
Open your MCP-compatible AI assistant and add a new custom MCP server.
When prompted:
Paste your Loxo MCP Server URL.
Sign in with your Loxo account.
Complete the authentication process.
Once connected, your AI assistant will be able to securely access your Loxo account based on your existing user permissions.
Using the Loxo MCP
Once you've connected the Loxo MCP to your AI assistant, you can interact with your Loxo data using natural language. For example, you can ask your AI assistant to:
Search for candidates
Find open jobs
View candidate or company details
Create notes
Update records (based on your permissions)
Summarize pipelines or recruiting activity
A few examples of prompts:
Example | What your AI can do |
Candidate Search | "Find software engineers in Denver with Kubernetes experience." |
Job Lookup | "Show me open engineering jobs." |
Candidate Updates | "Add a note that I spoke with Sarah yesterday." |
Recruiting Workflow | "Summarize every candidate in this pipeline." |
Note: The Loxo MCP Connector is currently in beta. Some advanced permission and visibility settings available in the Loxo UI may not yet be fully reflected when accessing data through MCP. If your organization relies on advanced data partitioning or strict team-based visibility, contact Support with any additional questions.
Troubleshooting Limitations
As the Loxo MCP is currently in Beta, you may encounter some known limitations or error responses while using certain tools and endpoints. The table below explains the most common errors, what they typically indicate and the recommended next steps.
Error | What It Likely Means | What To Do |
HTTP 500 (HTML error page) | Some endpoints return an HTML error page instead of structured JSON when they fail. | Capture the request ID from the response and report it. The error body cannot be reliably parsed. |
HTTP 400 Bad Request | Known limitation of the Beta | This is a known issue on certain write endpoints. The current best use case for the Loxo MCP is for reading data and providing analysis. |
HTTP 403 | Webhook endpoints only support API key authentication. OAuth (used by MCP) is not supported. | Manage webhooks using API key authentication outside of the MCP connection. |
Empty results from | The endpoint exists but does not currently return task data, even when tasks are visible in the UI. | View task information in the Loxo UI instead. Treat this as a known API limitation. |
As always, reach out to Support if you have additional questions!