ArcadeAI/arcade-mcp
MCP Server Framework and Tool Development library for building custom capabilities into agents. observed · 2026-09-03
Health v2 · maintenance only
69/100
- Activity 100
- Release rhythm 35
- Longevity 61
Flags: no_releases
How is this computed?
round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.
- gap_med: n/a
- age_days: 859
- days_rel: n/a
- days_push: 0
- n_releases_24m: 0
Adoption not part of the score
1016 stars · 113 forks observed · 2026-09-03
What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded
Arcade MCP is an open-source Python framework for building Model Context Protocol (MCP) servers and the tools that run inside them, with a decorator API covering the full MCP spec. It provides authorized tool calling with managed OAuth and secrets, tool-call evaluation against real LLMs, and one-command deployment.
Use cases
- build a custom MCP server in python
- create tools for claude desktop or cursor
- add oauth-authorized tool calling to ai agents
- expose internal apis as mcp tools
- test llm tool-call accuracy
- connect custom tools to langchain or crewai agents
- host mcp servers for my organization
When to choose
- you need MCP tools not available in prebuilt catalogs, like internal APIs or custom OAuth providers
- you want secrets and OAuth tokens hidden from the LLM and client
- you need vendor-neutral MCP servers that work with any client or agent framework
- you want to evaluate tool-call accuracy against real LLMs
When to avoid
- you only need existing prebuilt tool integrations and never custom tools
- you want a non-Python MCP framework
- you need a full agent orchestration framework rather than a tool/server layer
Facets
framework · maturity active
mcp agent-framework auth sdk api-framework testing deployment artificial-intelligence large-language-models developer-tools apis python cross-platform self-hosted cloud mcp-server model-context-protocol tool-calling oauth llm-tools claude decorator-api tool-evaluation ai-agents
10 sources
- readme: https://github.com/ArcadeAI/arcade-mcp · fetched 2026-09-03 · ff1c2db4655b
- homepage: https://docs.arcade.dev · fetched 2026-08-29 · 05f7a028c04d
- site_page: https://docs.arcade.dev/en/get-started/about-arcade · fetched 2026-08-29 · 50762a9f8eae
- site_page: https://docs.arcade.dev/en/get-started/setup/connect-arcade-docs · fetched 2026-08-29 · 68a3c0de1cbb
- site_page: https://docs.arcade.dev/en/get-started/quickstarts/call-tool-agent · fetched 2026-08-29 · 0752faa0f191
- site_page: https://docs.arcade.dev/en/get-started/quickstarts/call-tool-client · fetched 2026-08-29 · 2db4ed112b7a
- site_page: https://docs.arcade.dev/en/get-started/quickstarts/mcp-server-quickstart · fetched 2026-08-29 · 0a054b5e5fe8
- site_page: https://docs.arcade.dev/operate/quickstart · fetched 2026-08-29 · c92a1661b932
- site_page: https://docs.arcade.dev/get-started/quickstarts/call-tool-agent · fetched 2026-08-29 · 172ccdd61e5a
- site_page: https://docs.arcade.dev/get-started/quickstarts/mcp-server-quickstart · fetched 2026-08-29 · bf8e1898cb1a
Member repositories
| Repository | Role | Health v2 |
|---|---|---|
| ArcadeAI/arcade-mcp | main | 69 |
For agents
markdown · JSON · MCP: product_card(name="ArcadeAI/arcade-mcp")
Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem