MCP App Store
by KeenEthics
Inngest icon

Inngest MCP Server for Claude and ChatGPT

By Inngest, Inc.

Overview

Connect Claude to your Inngest Cloud account to inspect deployed apps and functions, diagnose failed or slow runs, and manage durable workflows. Explore environments, inspect run traces and event data, and query execution history with Insights. Send events, invoke functions, rerun or cancel executions, and sync applications. Sign in with your Inngest account through OAuth. Available tools and data depend on your account’s permissions and enabled features.

How to connect

  1. Claude: Open the Inngest listing in the Claude connectors directory and choose Connect. Claude listing Inngest exposes 54 tools here.
  2. ChatGPT: Open the Inngest listing in the ChatGPT app directory and choose Connect. ChatGPT listing Inngest exposes 76 tools here.

Tools

cancel_build

ChatGPT
Cancel an unfinished build

cancel_deploy

ChatGPT
Cancel an unfinished deploy

cancel_run

ChatGPT
Cancels an in-progress function run

commit_build

ChatGPT
Commit an upload and start the build

commit_deploy

ChatGPT
Commit an upload and start the deploy

create_build

ChatGPT
Upload one linux/amd64 docker-save tar or tar.gz to upload_url in a PUT request with every header in upload_headers, then commit the upload to start conversion. Requires image-build access.

create_cancellation

ChatGPT
Cancels a function's runs queued between from and until, inclusive, with millisecond precision. until may be up to now plus five seconds. There is no cap on range width; a from far in the past includes all queued runs through until. Includes runs that have never started, even when the function is paused. Running runs stop before their next step; sleeping or waiting runs are cancelled when they next resume. A step already executing may finish. The optional if expression filters the triggering event. Returns 201 when created; cancellation happens asynchronously. Not idempotent: each request creates another cancellation.

create_deploy

ChatGPT
Provide config from the selected apps entry in inngest.toml with exactly one of serve or connect. Memory defaults to 2048 MiB, CPU to 1 vCPU, and connect replicas to 1. Serve path defaults to /api/inngest and an omitted port is discovered from the image's exposed ports at launch. Omitted serve_config settings use node defaults. Upload one linux/amd64 docker-save tar or tar.gz to upload_url in a PUT request with every header in upload_headers, then commit the upload to build a function artifact. Runtime ENV, ENTRYPOINT, and CMD come from the image. Requires image-build access.

create_env

ChatGPT
Create custom environment

create_replay

ChatGPT
Creates an asynchronous replay for matching runs of one function. Cloud bounds have millisecond precision: from rounds down; until includes the end of its millisecond. Dev server bounds are inclusive at stored timestamp precision. Dev server replays run in memory and are lost on restart. Requests are not idempotent: retrying creates another replay. In Inngest Cloud, replays are listed on the function's Replays page in the dashboard. Scheduling is best effort and does not preserve event order.

create_sandbox

ChatGPT
Create a sandbox

create_sandbox_snapshot

ChatGPT
Create a sandbox snapshot

create_score

ChatGPT
Submits one or more named scores for a function run, or for specific steps when step IDs are provided. Scores are recorded as run metadata and aggregated for experiments.

create_webhook

ChatGPT
Creates a new webhook for receiving events

delete_sandbox_snapshot

ChatGPT
Delete a sandbox snapshot

destroy_sandbox

ChatGPT
Destroy a sandbox

exec_sandbox

ChatGPT
Execute a command in a sandbox

fetch_account

ChatGPT
Returns the account for the authenticated user

fetch_account_event_keys

ChatGPT
Lists event keys for a specific account, optionally filtered by environment. If the environment header isn't specified, this returns the default production environment's keys.

fetch_account_signing_keys

ChatGPT
Lists signing keys for a specific account, optionally filtered by environment. If the environment header isn't specified, this returns the default production environment's keys.

get_app

ChatGPT
Fetches details for a single app, including sync metadata and function count

get_app_sync

ChatGPT
Gets an app sync. Cloud keeps syncs until the app is deleted, with no separate retention window. The dev server does not support sync history and returns 501 for both list and get. Use GetApp latestSync for the dev server's current state.

get_apps

ChatGPT
Lists active apps in the authenticated environment. Set archived to true to list archived apps instead.

get_build

ChatGPT
Get a build

get_deploy

ChatGPT
Get a deploy

get_event

ChatGPT
Gets an event by its internal ID. Internal inngest/ events are not returned. Payloads are omitted unless include=payload. The dev server returns events stored in its database, which is in-memory by default and lost on restart. Cloud returns the same events as the dashboard and v1 API, subject to the plan's event retention.

get_event_runs

ChatGPT
Lists function runs triggered by a specific event

get_experiment

ChatGPT
Fetches variant run counts and score aggregates for one experiment on one function.

get_function

ChatGPT
Fetches function configuration and status details for a function within an app

get_run

ChatGPT
Fetches the canonical run summary for a single function run

get_run_trace

ChatGPT
Fetches the trace tree for a single function run

get_sandbox

ChatGPT
Get a sandbox

get_sandbox_process

ChatGPT
Get a sandbox process

get_sandbox_process_output

ChatGPT
Get sandbox process output

get_sandbox_snapshot

ChatGPT
Get a sandbox snapshot

health

ChatGPT
Returns the health status of the API service

invoke_function

ChatGPT
Invokes a function, executing the function either asynchronously or synchronously based on the mode parameter in the request body

list_app_syncs

ChatGPT
Lists app syncs newest first. Cloud keeps syncs until the app is deleted, with no separate retention window. The dev server does not support sync history and returns 501 for both list and get. Use GetApp latestSync for the dev server's current state.

list_builds

ChatGPT
List recent workspace builds

list_deploys

ChatGPT
List recent workspace deploys

list_envs

ChatGPT
List of all custom environments.

list_events

ChatGPT
Lists events in descending internal-ID order, excluding internal inngest/ events. Time filters use received time; client-seeded IDs can differ from received-time order. Payloads are omitted unless include=payload. The dev server returns events stored in its database, which is in-memory by default and lost on restart. Cloud returns the same events as the dashboard and v1 API, subject to the plan's event retention.

list_experiments

ChatGPT
Lists observed experiments in the authenticated environment. Use the nested app and function route to scope results to one function.

list_function_runs

ChatGPT
Lists runs for one function in the authenticated environment

list_functions

ChatGPT
Lists function configuration and status details for an app

list_insights_event_schemas

ChatGPT
Returns a paginated list of event type schemas, where each schema describes the shape of a specific event's data as nested JSON.

list_insights_tables

ChatGPT
Lists the available tables that can be queried via the Insights query endpoint.

list_runs

ChatGPT
Lists runs in the authenticated environment, optionally filtered by app and function IDs

list_sandbox_processes

ChatGPT
List sandbox processes

list_sandbox_snapshots

ChatGPT
List sandbox snapshots

list_sandboxes

ChatGPT
List sandboxes

list_session_keys

ChatGPT
Lists session keys observed in the authenticated environment.

list_session_runs

ChatGPT
Lists runs associated with one session ID in the authenticated environment.

list_sessions

ChatGPT
Lists session IDs for one session key in the authenticated environment.

list_webhooks

ChatGPT
Lists all webhooks in the specified environment

patch_env

ChatGPT
Partially updates an environment. Only the provided fields will be modified.

pause_function

ChatGPT
Pauses a function. Idempotent: pausing a paused function returns 200 and changes nothing. Send no request body. Runs already executing may continue. In Cloud, the pause can take up to a minute to apply, new runs are skipped while paused, and runs still unfinished after 7 days paused are cancelled, same as pausing in the dashboard. Wait for a pause or resume to finish before sending the opposite action: requests sent close together can be skipped or apply out of order. Confirm with GetFunction after the cache has refreshed and retry if needed. In the dev server, the pause applies right away, new runs stay queued, and it lasts until resumed or restarted. No bulk pause; call once per function.

pause_functions

ChatGPT
Pauses functions in one app. Idempotent. Accepts up to 500 unique IDs, at most 256 bytes each, in a body of at most 256 KiB of JSON. Split larger sets into multiple requests. Cloud limits bulk calls to a lower rate (about 1 per 5 seconds on most plans), so quick retries can get a 429. Returns one result per ID in request order; failures do not stop the rest. An unknown app returns 200 with not_found for every ID, so check each result. The path app ID is used; any appId in the body is ignored. Running work may continue. In Cloud, pauses can take up to a minute to apply, new runs are skipped, and unfinished runs are cancelled after 7 days paused. In the dev server, pauses apply immediately, new runs stay queued, and pauses last until resumed or restarted. Wait for each pause or resume request to finish before sending the opposite action; confirm with GetFunction after the cache refreshes and retry if needed.

pause_sandbox

ChatGPT
Pause a sandbox

query_insights

ChatGPT
Query Insights using the provided SQL query.

query_insights_prompt

ChatGPT
Translates a natural language prompt into an Insights SQL query.

read_sandbox_file

ChatGPT
Read a sandbox file

rerun

ChatGPT
Creates a new run using the original run's triggering event data

resume_function

ChatGPT
Resumes a paused function. Idempotent: resuming a function that is not paused returns 200 with isPaused: false. The dev server still wakes the queue so retries can recover a failed wake. Send no request body. In Cloud, the response reflects the requested state; the resume can take up to a minute to apply. A future scheduled pause returns 409 and is left unchanged. New runs skipped while paused are not recreated. Wait for a pause or resume to finish before sending the opposite action: requests sent close together can be skipped or apply out of order. Confirm with GetFunction after the cache has refreshed and retry if needed. In the dev server, resume clears the pause after waking the queue; a failed wake leaves the pause in place. Queued runs can then start. No bulk resume; call once per function.

resume_functions

ChatGPT
Resumes functions in one app. Idempotent. Accepts up to 500 unique IDs, at most 256 bytes each, in a body of at most 256 KiB of JSON. Split larger sets into multiple requests. Cloud limits bulk calls to a lower rate (about 1 per 5 seconds on most plans), so quick retries can get a 429. Returns one result per ID in request order; failures do not stop the rest. An unknown app returns 200 with not_found for every ID, so check each result. The path app ID is used; any appId in the body is ignored. In Cloud, the response reflects the requested state; resumes can take up to a minute to apply. Future scheduled pauses return per-function errors and stay unchanged. Skipped runs are not recreated. In the dev server, each queue is woken before its pause is cleared; a failed wake leaves that function paused. Retries wake queues again. Wait for each pause or resume request to finish before sending the opposite action; confirm with GetFunction after the cache refreshes and retry if needed.

resume_sandbox

ChatGPT
Resume a sandbox

send_event

ChatGPT
Sends one event for testing and debugging. The request accepts the name, data, user, id, and ts fields; send batches and other Event API fields through an Inngest SDK or the Event API. This endpoint uses REST API rate limits and is not intended for high-volume ingestion. Use the returned event ID with the event runs endpoint to inspect any function runs triggered by the event.

signal_sandbox_process

ChatGPT
Signal a sandbox process

start_sandbox_process

ChatGPT
Start a sandbox process

stream_sandbox_logs

ChatGPT
Stream sandbox logs

stream_sandbox_process_output

ChatGPT
Stream sandbox process output

submit_feedback

ChatGPT
Sends user-requested product feedback to Inngest support. Cloud only. Submit only feedback the user has asked to share; do not include credentials, secrets, or run payloads. Creates a support ticket, so retries may create duplicates. Requires feedback:write:create. No environment selection is required.

sync_app

ChatGPT
Sync an app at the provided URL.

update_webhook

ChatGPT
Updates a webhook's name or transform. Omitted fields are left unchanged.

wait_sandbox_process

ChatGPT
Wait for a sandbox process

write_sandbox_file

ChatGPT
Write a sandbox file

cancel_run

Claude

create_env

Claude

create_sandbox

Claude

create_sandbox_snapshot

Claude

create_score

Claude

create_webhook

Claude

delete_sandbox_snapshot

Claude

destroy_sandbox

Claude

exec_sandbox

Claude

fetch_account

Claude

fetch_account_event_keys

Claude

fetch_account_signing_keys

Claude

get_app

Claude

get_apps

Claude

get_event_runs

Claude

get_experiment

Claude

get_function

Claude

get_run

Claude

get_run_trace

Claude

get_sandbox

Claude

get_sandbox_process

Claude

get_sandbox_process_output

Claude

get_sandbox_snapshot

Claude

health

Claude

invoke_function

Claude

list_envs

Claude

list_experiments

Claude

list_function_runs

Claude

list_functions

Claude

list_insights_event_schemas

Claude

list_insights_tables

Claude

list_runs

Claude

list_sandbox_processes

Claude

list_sandbox_snapshots

Claude

list_sandboxes

Claude

list_session_keys

Claude

list_session_runs

Claude

list_sessions

Claude

list_webhooks

Claude

patch_env

Claude

pause_sandbox

Claude

query_insights

Claude

query_insights_prompt

Claude

read_sandbox_file

Claude

rerun

Claude

resume_sandbox

Claude

send_event

Claude

signal_sandbox_process

Claude

start_sandbox_process

Claude

stream_sandbox_logs

Claude

stream_sandbox_process_output

Claude

sync_app

Claude

wait_sandbox_process

Claude

write_sandbox_file

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/inngest"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/inngest)