Beeceptor MCP. Mock APIs and Debug Webhooks with Your AI Agent

You're gonna build, manage, and debug mock APIs and HTTP proxies with your AI client using this server. It lets you simulate complex backends without writing a line of actual code. You can control everything from the initial network setup to deep session state management.

When you start up, if you need a clean slate for testing, you've got tools like delete_all_rules and delete_requests right here—they clear out every single mock rule or historical request log entry on the endpoint. You can also wipe out stored session data using delete_state, or just remove one key-value pair with delete_state_item.

If you uploaded a big OpenAPI/Swagger file, you can scrap it entirely with delete_spec.

To define how your mock API behaves, you've got total control over the rules. You start by uploading industry-standard specs using upload_spec, which ingests and validates the data to generate a full set of initial mock endpoints. If you need to change those rules later, you can create a single rule with create_rule or overwrite everything an existing rule uses with update_rule_full.

For minor tweaks—say, changing just the status code or the response body without touching other settings—you use update_rule_partial. You can get a read on what rules are active using list_rules, and you can fetch all the details for one specific rule using get_rule.

Dealing with multiple rules? You can reorder them with reorder_rules, which is crucial when your routing logic depends on sequence. If you need to manage those specs, you can get their full metadata structure by calling get_spec_details. Sometimes you just want to check what rules are already set up; that's where list_rules comes in handy.

When it comes to webhooks and debugging incoming traffic, this server is your best friend. You can list all recent historical requests with list_requests, getting a paginated overview of everything that hit the mock API. For deep dives, you retrieve every single detail—headers, body, everything—for one specific request using get_request.

If you're debugging stateful calls, you manage data persistence by setting or updating key-value pairs in bulk with upsert_state, or reading a stored value for a single key using get_state_item; you can also list all current keys and values using list_state.

To get the whole picture of your setup, you check the global network settings with get_settings. Here you manage access control by modifying CORS headers using update_settings, or adding security layers by attaching an mTLS certificate via add_certificate; if that cert needs removing, you use delete_certificate.

Need to send data back in a mock response? You can upload binary files like images or documents with upload_blob so they show up when your client calls the endpoint. When testing complex scenarios, you also have tools to manage state history: you get status updates for long jobs using get_job_status, and if you're doing a bulk cleanup of stored data items, you use delete_state.

You can check what certificates are currently configured by running list_certificates.

Every action is logged. Besides the general request history, you get full visibility into how your endpoint operates by getting the current configuration and details for any single rule with get_rule, or simply checking which rules exist using list_rules. If you want to see what data was stored before a test run, you can list all state keys using list_state.