n8n Multi Instance skill
Target multiple n8n instances explicitly and recover from ambiguous MCP routing.
Community · Romuald Członkowski| Author | Romuald Członkowski |
|---|---|
| Licence | MIT |
| Source | czlonkowski/n8n-skills · SKILL.md on GitHub |
| Project website | docs.n8n.io |
| Works in | Claude Code · Codex, Gemini CLI, Cursor, Copilot |
| Read from the source on |
n8n Multi Instance skill: Target multiple n8n instances explicitly and recover from ambiguous MCP routing.
Discover available instance names, name the target on each supported call and verify credential-write destinations.
From the skill's SKILL.md: # Working with multiple n8n instances over MCP
What it does
- Discover available instance names, name the target on each supported call and verify credential-write destinations.
What you get
- Instance-aware call plans, routing checks and recovery steps for wrong-target errors.
When to use it
- List the instances and plan this change for staging only.
When not to use it
- Treating instructions as a guarantee that a remote service, validation or deployment has succeeded.
What it needs
- A compatible n8n-mcp server and actual workflow data. Live management needs an authorized n8n instance; connected services and hosting can have separate costs.
Files in the skill
| File or folder | What it is for |
|---|---|
SKILL.md | Entry instructions and task-specific workflow. |
Licence as found in the repository on 2026-11-17: MIT.
Example prompts
Type one of these after installing; naming the skill makes sure it is used.
- List the instances and plan this change for staging only.
- Explain INSTANCE_AMBIGUOUS without recreating missing objects.
- Check the exact target before changing credentials.
How to install n8n-multi-instance
Prepare the workflow
The website link is n8n product documentation, not the community MCP server installer. Install n8n-mcp separately; hosting and connected services may cost money. Review plugin hooks before installation.
Claude Code plugin hooks are not installed by copying a skill folder to another agent. Check the live tool schema and review writes before execution.
A skill is a folder with a SKILL.md file. You install it once; the agent reads its short description at the start of each session and loads the full instructions only when a request matches. Pick the place you use:
Claude Code: from the plugin marketplace
- Start Claude Code in a terminal with
claude. - Add the marketplace (once per computer):
/plugin marketplace add czlonkowski/n8n-skills - Install the plugin. Claude Code opens its details first: check what it will add (skills, hooks, MCP servers), then choose whether it is for you, for everyone on this repository, or for you in this repository only.
/plugin install n8n-mcp-skills@n8n-mcp-skills - Type / and look for
/n8n-mcp-skills:n8n-multi-instance, or simply ask for the task.
From your shell instead of a session:
claude plugin marketplace add czlonkowski/n8n-skills
claude plugin install n8n-mcp-skills@n8n-mcp-skillsThe plugin installs as a whole, so other skills come with it, including n8n-subworkflows, n8n-self-hosting, n8n-binary-and-data, n8n-node-configuration, n8n-code-javascript, n8n-workflow-patterns, using-n8n-mcp-skills, n8n-code-python, n8n-agents, n8n-expression-syntax, n8n-error-handling, n8n-validation-expert, n8n-mcp-tools-expert, n8n-code-tool.
Claude Code: copy the folder into your skills folder
Put the skill's folder in ~/.claude/skills/ to use it in every project on this computer, or in .claude/skills/ inside a repository to share it with everyone who works there (commit it).
git clone --depth 1 https://github.com/czlonkowski/n8n-skills.git
mkdir -p ~/.claude/skills
cp -R n8n-skills/skills/n8n-multi-instance ~/.claude/skills/Claude Code notices new or changed skills in these folders during the session. If the skills folder did not exist when the session started, run /reload-skills once. Then type /n8n-multi-instance or just describe the task.
On Windows, run the commands in Git Bash or WSL, or copy the folder into the same place with File Explorer.
Codex, Gemini CLI, Cursor and GitHub Copilot
These agents read the same SKILL.md folders. The one folder all four accept inside a project is .agents/skills/, so this copies the skill there:
git clone --depth 1 https://github.com/czlonkowski/n8n-skills.git
mkdir -p .agents/skills
cp -R n8n-skills/skills/n8n-multi-instance .agents/skills/| Agent | Project folder | Your user folder | Notes |
|---|---|---|---|
| OpenAI Codex | .agents/skills/ | ~/.agents/skills/ | Type $ or run /skills to pick a skill; Codex also chooses one when your task matches. |
| Gemini CLI | .agents/skills/.gemini/skills/ | ~/.agents/skills/~/.gemini/skills/ | Or install from Git with gemini skills install <git URL> --path <folder> --consent. Google replaced Gemini CLI with Antigravity CLI for free-tier and Google One users on 18 June 2026; Antigravity uses .agents/skills/ in a project and ~/.gemini/config/skills/ for you. |
| Cursor | .agents/skills/.cursor/skills/ | ~/.agents/skills/~/.cursor/skills/ | Type / in Agent chat to pick a skill. Cursor also reads Claude Code's .claude/skills/ folders. |
| GitHub Copilot | .github/skills/.agents/skills/.claude/skills/ | ~/.copilot/skills/~/.agents/skills/ | Works in Copilot CLI, the cloud agent, and agent mode in VS Code, Visual Studio and JetBrains (IDE support depends on the version). Or run gh skill install OWNER/REPOSITORY SKILL (public preview, GitHub CLI 2.90.0 or later). |
Sources for these steps
Checked against each vendor's documentation on 2026-11-14:
Discover before writing
Use the exact available name, not a guessed URL or ID. An unexpected NOT_FOUND can indicate wrong routing; verify the instance before recreating anything.
Credential boundaries
Name and confirm the destination for sensitive writes. Ambiguity errors fail closed; choose the intended instance rather than bypassing the error blindly.
Questions about n8n-multi-instance
Does this website execute the skill?
No. This is a source-based installation guide. Your agent runs the skill in its own configured environment.
Are full cloud workflows verified here?
No. Generic cloud upload and API execution steps are omitted; source-specific local requirements and host capabilities still apply.
Does switching affect other chats?
It can. Hosted connectors may share a default across chats; explicit instance arguments route one call without changing the shared default.
What if instance is not in the tool schema?
The server may predate named-call routing. Verify capabilities first; switch separately before dependent work, without assuming parallel call order.
More AI skills
Related skills
n8n Self Hosting skill
Community · Romuald CzłonkowskiDeploy and maintain Docker-based n8n with Caddy HTTPS, private services and verified backups.n8n Code JavaScript skill
Community · Romuald CzłonkowskiWrite or debug n8n JavaScript Code nodes with mode-aware data access and returns.n8n Code Python skill
Community · Romuald CzłonkowskiWrite native Python for n8n 2.x Code nodes with explicit sandbox and mode limits.n8n Code Tool skill
Community · Romuald CzłonkowskiWrite agent-callable n8n Custom Code Tools with query input and string output.n8n Agents skill
Community · Romuald CzłonkowskiDesign n8n agents with deliberate tools, memory, structured output and human review.n8n Node Configuration skill
Community · Romuald CzłonkowskiConfigure n8n nodes from live schemas, operations and property dependencies.