MCP servers for Claude Code · Automation / agent builders
Best automation / agent builders MCP server for Claude Code, tested
In ToolPassed's hands-on tests in Claude Code, n8n-mcp is our provisional candidate for automation / agent builders: it passed install, connect, tools and an agent task on 2026-10-21. We ran 3 automation / agent builders MCP servers in Claude Code; 3 passed every step.
Last tested · runs from 2026-10-16 to 2026-10-21 · one run per server · How we test
Results in Claude Code
| Server | Claude Code |
|---|---|
| n8n-mcp Automation / agent builders | Passed 2026-10-21 · install, connect, tools, agent task |
| n8n instance-level MCP Automation / agent builders | Passed 2026-10-16 · install, connect, tools, agent task |
| Langflow MCP server Automation / agent builders | Passed 2026-10-16 · install, connect, tools, agent task |
Dated results
- n8n-mcp passed in Claude Code on 2026-10-21 (Claude Code 2.1.296, n8n-mcp 2.94.3, connected in 6.2s). Full result
- n8n instance-level MCP passed in Claude Code on 2026-10-16 (Claude Code 2.1.296, server reported v1.2.0, connect time not recorded). Full result
- Langflow MCP server passed in Claude Code on 2026-10-16 (Claude Code 2.1.296, server reported v1.30.0, connect time not recorded). Full result
Why n8n-mcp is the candidate
n8n-mcp passed every step in Claude Code. It is the group’s site-wide candidate (judged 2026-10-21): Searched n8n's node library for 'slack' and returned the Slack node in both Claude Code and Cursor (2 calls in total), with nothing but the published snippet: no n8n instance, account or key. It also passed in ChatGPT (Codex) on Oct 15. Runner-up Langflow ran our fixture flow and returned the fresh code in both clients in 2 calls in total, but it needs a running Langflow server and an API key. n8n's own instance-level MCP passed both tasks (node search and running a fixture workflow) in both clients too, but needs a running n8n instance with an access token and took 8 calls on the workflow task (4 per client) against Langflow's 2. Confirmed Oct 21, 2026: a second run in Claude Code and Cursor passed again (n8n-mcp 2.94.3).
Provisional: based on one dated run per app (no repeat runs yet), so treat it as a candidate, not a final pick. How we test.
Same servers in all 4 apps
| Server | Claude Code | Cursor | Claude Desktop | ChatGPT (Codex) |
|---|---|---|---|---|
| n8n-mcp Automation / agent builders | Passed 2026-10-21 · install, connect, tools, agent task | Passed 2026-10-21 · install, connect, tools, agent task | Passed 2026-10-19 · install, connect, tools, agent task | Passed 2026-10-15 · install, connect, tools, agent task |
| n8n instance-level MCP Automation / agent builders | Passed 2026-10-16 · install, connect, tools, agent task | Passed 2026-10-16 · install, connect, tools, agent task | Not tested yet | Not tested yet |
| Langflow MCP server Automation / agent builders | Passed 2026-10-16 · install, connect, tools, agent task | Passed 2026-10-16 · install, connect, tools, agent task | Not tested yet | Not tested yet |
What failed or is partial in Claude Code
None of these servers failed in Claude Code.
Install in Claude Code
The exact snippet we tested is on the n8n-mcp page (pick the Claude Code tab). Not connecting? Claude Code MCP troubleshooting.
Not tested in Claude Code yet
ARTEMIS (Google), Home Assistant MCP (ha-mcp), agent-device (Callstack), Open Computer Use, Aident Loadout, API.market MCP Gateway. Listed from their source; no result in Claude Code yet.
Questions
What is the best automation / agent builders MCP server for Claude Code?
In ToolPassed's hands-on tests in Claude Code, n8n-mcp is our provisional candidate for automation / agent builders: it passed install, connect, tools and an agent task on 2026-10-21. We ran 3 automation / agent builders MCP servers in Claude Code; 3 passed every step.
Does n8n-mcp work with Claude Code?
n8n-mcp passed in Claude Code on 2026-10-21 (Claude Code 2.1.296, n8n-mcp 2.94.3, connected in 6.2s).
More: all MCP servers tested in Claude Code · Automation / agent builders in every app · How we test