# UsefulSoftwareCo/executor

The missing integration layer for AI agents. Let them call any OpenAPI / MCP / GraphQL / custom js functions in secure environment.

Repository: https://github.com/UsefulSoftwareCo/executor
Canonical: https://ross.abutalabs.com/products/executor
Homepage: https://executor.sh
Language: TypeScript
License: MIT
License Family: permissive
Last push: 2026-08-26T22:24:54+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 14
- inputs: {"age_days": 207, "days_push": 7, "days_rel": 7, "gap_med": 0.0, "n_releases_24m": 81}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3200, forks 256 (observed 2026-08-28T04:07:48.253054+00:00)

## What it is
Executor is an open-source integration layer and MCP gateway for AI agents, letting you configure MCP servers, OpenAPI specs, GraphQL APIs, and custom TypeScript functions once with authentication and per-tool policies. Any MCP-compatible agent (Claude Code, Cursor, Codex) then shares one tool catalog through a single endpoint, with tools loaded on demand to keep context windows small.

## Use cases
- connect ai agents to external apis and tools
- share mcp integrations across multiple agents
- manage api credentials for ai agents in one place
- reduce token usage from too many mcp tools
- set approval policies for agent tool calls
- self-host an mcp gateway for my team
- expose openapi and graphql apis to claude or cursor

## When to choose
- you use multiple MCP-compatible agents and keep duplicating integrations and API keys
- you need per-tool governance like allow, approval-gated, or blocked policies
- your agent's context window is bloated by hundreds of tool schemas
- you want a single self-hosted container with no external database to run

## When to avoid
- your agent is not MCP-compatible
- you only use one agent with a handful of integrations and don't need a shared catalog
- you need deep custom logic beyond what JSON-schema-described tools or sandboxed TypeScript can express

## Facets
- artifact type: service
- maturity: active
- function: mcp, api-framework, middleware, auth, security, plugin-system, http-server, cli, developer-tools
- domain: large-language-models, apis, developer-tools, self-hosted
- platform: self-hosted, cloud, cli, cross-platform, windows
- tags: mcp-gateway, tool-catalog, openapi, graphql, integration-layer, agent-tools, context-efficiency, per-tool-policies, sandboxed-execution, ai-agents, automation, nodejs, typescript, docker, desktop, macos, linux

## Member repositories
- UsefulSoftwareCo/executor (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:48.253054+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:24:49.982356+00:00, confidence not recorded.
  - readme: https://github.com/UsefulSoftwareCo/executor (fetched 2026-08-28T04:07:48.253054+00:00, sha b8659dfc301b)
  - homepage: https://executor.sh (fetched 2026-08-29T09:38:52.317396+00:00, sha 92f9063bf9f3)
  - site_page: https://executor.sh/docs (fetched 2026-08-29T09:38:52.320365+00:00, sha 53d9c43faa2e)
  - site_page: https://executor.sh/docs/hosted/docker (fetched 2026-08-29T09:38:52.322559+00:00, sha eb0de224aa4e)
- Data as of 2026-08-30T08:39:29.467469+00:00.