n8n-flow-bridge-mcp
Enables git-based version control for n8n workflows with automatic credential remapping, allowing AI tools to pull, push, and manage workflows across environments.
README
n8n Flow Bridge
Git-based version control for n8n workflows with automatic credential remapping — stop rewiring nodes every import.
The problem
Exporting/importing n8n workflows breaks credential bindings because node references point to internal instance-specific ids, not names. Every environment switch (dev → prod, or teammate → teammate) means manually reopening every credentialed node and rewiring it by hand. This tool fixes that.
What it does
- Pulls workflows from n8n into clean, diffable git files
- Auto-rewrites credential references to stable symbolic keys
(
cred:<type>:<slug>) - Auto-remaps symbolic keys back to real local credential ids on push
- Detects and blocks on missing credential bindings before they break a workflow silently
Quick start
No install step — works the same on Windows, macOS, and Linux:
export N8N_BASE_URL=http://localhost:5678
export N8N_API_KEY=...
npx n8n-flow-bridge-mcp pull --workflow 123
npx n8n-flow-bridge-mcp bind cred:slackApi:team-bot
npx n8n-flow-bridge-mcp push --workflow 123
Prefer a shorter command? npm install -g n8n-flow-bridge-mcp once, then use
bridge directly (bridge pull --workflow 123, etc.) instead of
npx n8n-flow-bridge-mcp ....
Developing on this repo instead of the published package
git clone https://github.com/dorkian/n8n-flow-bridge-mcp.git
cd n8n-flow-bridge-mcp
npm install
npm run build
npm link # makes the local build available as `bridge`
Running separate dev and prod n8n instances from one checkout? Every command
accepts -e, --env-dir <path> (and -d, --dir <path> for workflows/
itself, rarely needed) to say which environment's .env and
credentials.map.json to use — point it at any two folders you like, e.g.:
bridge push --workflow 123 --env-dir ./environments/dev
bridge push --workflow 123 --env-dir ./environments/prod
workflows/ stays a single shared, git-tracked folder; only .env and
credentials.map.json differ per environment. Register bridge mcp --env-dir <path> twice (once per environment) to get the same separation
through MCP tools.
Before / after
Raw n8n export (instance-specific, breaks on import elsewhere):
{
"name": "Post Release Alert",
"type": "n8n-nodes-base.slack",
"credentials": {
"slackApi": { "id": "17", "name": "Team Bot" }
}
}
After bridge pull (portable, committed to git):
{
"name": "Post Release Alert",
"type": "n8n-nodes-base.slack",
"credentials": {
"slackApi": { "id": "cred:slackApi:team-bot", "name": "Team Bot" }
}
}
bridge push remaps cred:slackApi:team-bot back to whatever the real
local credential id is on the target machine — see
docs/credential-remap.md for the full
walkthrough.
CLI commands
| Command | What it does |
|---|---|
bridge pull --workflow <id> / --all |
Fetch, sanitize, and save workflow(s) to workflows/ |
bridge push --workflow <id|file|name> / --all |
Remap credentials and push workflow(s) to n8n |
bridge bind <cred-key> |
Bind a symbolic credential key to a local credential id |
bridge bind --auto |
Auto-bind all unbound keys by exact type + name match |
bridge status |
Show credential drift and remote sync status |
bridge mcp |
Run the MCP server (stdio) for Claude Code / Cursor |
All of the above accept -e, --env-dir <path> (which .env +
credentials.map.json to use) and -d, --dir <path> (where workflows/
lives, defaults to cwd, rarely needs overriding).
Configuration
Set N8N_BASE_URL and N8N_API_KEY (env vars or a .env file, optionally
loaded from --env-dir).
Want AI to build the workflow itself, not just version it?
This tool has no opinion about which n8n nodes solve your problem — pair it
with n8n-mcp for that (node
catalog, validation, natural-language workflow creation/editing on dev),
and use bridge to snapshot the result into git and promote it to prod.
See docs/pairing-with-n8n-mcp.md.
Architecture
bridge pull fetches a workflow from n8n, strips instance-specific
metadata, and rewrites each node's credential id to a portable symbolic key
(sanitize.ts) before writing it to workflows/ in a deterministically
ordered, diff-friendly format (diffFormat.ts). Each machine keeps a
git-ignored credentials.map.json binding those symbolic keys to its own
real credential ids (credentialMap.ts). bridge push substitutes the
real ids back in (remap.ts) — aborting first if any key is unbound — then
calls the n8n API. The CLI (src/cli/*) and the MCP server
(src/mcp/server.ts) both call the same core modules, so both surfaces stay
in sync. Full details in docs/architecture.md.
Roadmap
- [x] MCP server for Claude Code / Cursor
- [x] Multi-environment support from one checkout (
--env-dir) - [ ] Web UI for credential binding
- [ ] Multi-instance orchestration (sync dev → staging → prod in one command)
See docs/roadmap.md.
About me
Ashkan Dorkian — AI-native frontend & automation engineer.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.