Executor
One place where agents connect to the tools you already use; wire it up once and the whole
Executor is an MCP gateway that connects MCP servers, OpenAPI specs, and GraphQL APIs behind a single tool shape, so agents connect once and share integrations across clients. It addresses the problem of configuring the same accounts and credentials separately in each agent, while keeping policies consistent.
Executor derives policies from the source protocol (GET versus DELETE for OpenAPI, destructiveHint for MCP, mutations for GraphQL), runs calls in an isolated JavaScript sandbox with credentials attached host-side, and exposes workspace-shared connections. It runs as Cloud, desktop (Mac, Windows, Linux), CLI, or self-hosted Docker, and Cloud offers an HTTP API with user and workspace keys. It is open source.
It is aimed at individuals and teams using agent clients such as Claude Code, Cursor, and Codex. Cloud is free for up to three members with a monthly execution allowance and unlimited integrations; a Team plan is priced per member with a 14-day trial and unlimited executions; an Enterprise edition adds self-hosted or dedicated cloud deployment, SSO/SAML, SCIM, audit logs, and support.
12 alternatives to Executor
Ranked by how well each tool replaces Executor: shared features, audience, price and popularity.
- 60 out of 100 match$20/mo
An open-source standard for connecting AI applications to external systems
Covers 3 of 15 key features.
Open source58 out of 100 match—DataRobot: Unified Agent Workforce Platform for Enterprise.
Covers 5 of 15 key features.
Free plan57 out of 100 matchContact sales- 57 out of 100 matchFree
Cloud platform for running open-source machine learning models via API
Covers 3 of 15 key features.
Free planOpen source57 out of 100 matchUsage-based- 56 out of 100 matchFree
- 56 out of 100 matchContact sales
- 56 out of 100 matchUsage-based
- 56 out of 100 matchFree
The AI community building the future
Covers 1 of 15 key features and starts cheaper.
Free plan55 out of 100 match$9/mo- 55 out of 100 matchUsage-based
- 55 out of 100 matchUsage-based