n8n-multi-instance

Name: n8n-multi-instance
Author: czlonkowski

438installs983forks

How to install n8n-multi-instance?

czlonkowski/n8n-skills/n8n-multi-instance

$npx -y skills add czlonkowski/n8n-skills --skill n8n-multi-instance

Installs into the current project.

Summary

Use when an n8n-mcp account targets more than one n8n instance — i.e. the n8n_instances tool is available, the user mentions multiple n8n instances or environments (prod vs staging, several teams or clients), a workflow / datatable / credential / execution call returns an unexp

Files · 1

View on GitHub

SKILL.md

1	# Working with multiple n8n instances over MCP
2
3	When the `n8n_instances` tool is available, the user has multi-instance mode on: one MCP
4	connection can reach several n8n instances (e.g. `prod`, `staging`, or one per client/team).
5	Every other n8n tool (`n8n_get_workflow`, `n8n_list_workflows`, `n8n_update_partial_workflow`,
6	`n8n_manage_datatable`, `n8n_manage_credentials`, `n8n_executions`, `n8n_test_workflow`, …) runs
7	against whichever instance this session is currently targeting. There is no per-call instance
8	argument: you change the target only by switching. Target the wrong instance and a read returns the
9	wrong data and a write lands in the wrong place — usually with no error (the one exception is an
10	ambiguous credential write, which fails closed; see below). So target deliberately.
11
12	If the `n8n_instances` tool is not present, the account is single-instance: ignore this skill
13	and use the n8n tools directly.
14
15	## Golden rules
16
17	Six rules. Each prevents a class of silent misroute.
18
19	1. Discover first. Call `n8n_instances({mode:"list"})` before acting so you know the instance
20	names and which one is `current`.
21	2. Switch by name to your target before doing work on a non-default instance:
22	`n8n_instances({mode:"switch", name:"<instance name>"})`. The match is case-insensitive.
23	3. Switch in its own turn. Never put a `switch` and a dependent operation in the **same
24	parallel tool-call batch**. Calls in one batch have no guaranteed order, so the dependent call
25	can be resolved against the previous instance before the switch's session state is visible.
26	Switch, let it return, then operate.
27	4. Verify before high-stakes ops. Immediately before creating/updating/deleting credentials
28	(and before destructive workflow edits), confirm `current` is the instance you intend — primary
29	check is `n8n_instances({mode:"list"})`. The system fail-closes only the ambiguous credential
30	case (rule 6); an explicit switch to the wrong instance still writes there silently, so this
31	check is on you.
32	5. An unexpected `NOT_FOUND` is almost always a wrong-instance misroute, not a deletion. Don't
33	recreate the object. Re-check the current instance and retry (see Recovery).
34	6. *On `INSTANCE_AMBIGUOUS`, switch on this* session, then retry.** The system is refusing to
35	write a secret because this session never picked a target itself. Comply — run `switch` here to
36	confirm the instance, then retry the write. Don't work around it or retry blindly.
37
38	## Core workflow
39
40	```
41	1. n8n_instances({mode:"list"}) # see available[] + current + default
42	2. n8n_instances({mode:"switch", name:"prod"}) # bind THIS session to "prod"
43	→ returns { previous, current }; confirm current.name == "prod"
44	3. (do your work) n8n_list_workflows / n8n_get_workflow / n8n_manage_datatable / ...
45	4. Before a credential write or a delete:
46	n8n_instances({mode:"list"}) → re-confirm current, THEN n8n_manage_credentials({action:"create", ...})
47	```
48
49	To move to another instance, just `switch` again. The whole session follows the switch.
50
51	## The `n8n_instances` tool
52
53	Two modes (`mode` is required and enum-validated):
54
55	- `{mode:"list"}` → `{ current, default, available }`, no side effects.
56	- `current` and `default` are each one instance `{ id, name, url, isDefault }` (or `null`).
57	- `available` is every instance, each with an extra `isCurrent` boolean. Match by `name`;
58	never hard-code `id`.
59	- `{mode:"switch", name:"<name>"}` → `{ previous, current }`, and binds this session to the named
60	instance. `name` is case-insensitive.
61
62	### Error envelope (from the `n8n_instances` tool)
63
64	Every error returns `{ error: "<CODE>", message, … }`. The ones you'll actually hit:
65
66	\| Code \| When \| What to do \|
67	\|---\|---\|---\|
68	\| `UNKNOWN_INSTANCE` \| `name` matches no instance \| Pick a name from the `available` list in the error payload and retry. \|
69	\| `NAME_REQUIRED` \| `switch` with no `name` \| Re-call with a `name` (the error lists the valid ones in `available`). \|
70	\| `MULTI_INSTANCE_DISABLED` \| multi-instance mode is off \| Th

People who liked this also viewed

6 picks

Type

n8n-multi-instance

438installs983forks

View on GitHub

How to install n8n-multi-instance?

czlonkowski/n8n-skills/n8n-multi-instance

$npx -y skills add czlonkowski/n8n-skills --skill n8n-multi-instance

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Working with multiple n8n instances over MCP
2
3	When the `n8n_instances` tool is available, the user has multi-instance mode on: one MCP
4	connection can reach several n8n instances (e.g. `prod`, `staging`, or one per client/team).
5	Every other n8n tool (`n8n_get_workflow`, `n8n_list_workflows`, `n8n_update_partial_workflow`,
6	`n8n_manage_datatable`, `n8n_manage_credentials`, `n8n_executions`, `n8n_test_workflow`, …) runs
7	against whichever instance this session is currently targeting. There is no per-call instance
8	argument: you change the target only by switching. Target the wrong instance and a read returns the
9	wrong data and a write lands in the wrong place — usually with no error (the one exception is an
10	ambiguous credential write, which fails closed; see below). So target deliberately.
11
12	If the `n8n_instances` tool is not present, the account is single-instance: ignore this skill
13	and use the n8n tools directly.
14
15	## Golden rules
16
17	Six rules. Each prevents a class of silent misroute.
18
19	1. Discover first. Call `n8n_instances({mode:"list"})` before acting so you know the instance
20	names and which one is `current`.
21	2. Switch by name to your target before doing work on a non-default instance:
22	`n8n_instances({mode:"switch", name:"<instance name>"})`. The match is case-insensitive.
23	3. Switch in its own turn. Never put a `switch` and a dependent operation in the **same
24	parallel tool-call batch**. Calls in one batch have no guaranteed order, so the dependent call
25	can be resolved against the previous instance before the switch's session state is visible.
26	Switch, let it return, then operate.
27	4. Verify before high-stakes ops. Immediately before creating/updating/deleting credentials
28	(and before destructive workflow edits), confirm `current` is the instance you intend — primary
29	check is `n8n_instances({mode:"list"})`. The system fail-closes only the ambiguous credential
30	case (rule 6); an explicit switch to the wrong instance still writes there silently, so this
31	check is on you.
32	5. An unexpected `NOT_FOUND` is almost always a wrong-instance misroute, not a deletion. Don't
33	recreate the object. Re-check the current instance and retry (see Recovery).
34	6. *On `INSTANCE_AMBIGUOUS`, switch on this* session, then retry.** The system is refusing to
35	write a secret because this session never picked a target itself. Comply — run `switch` here to
36	confirm the instance, then retry the write. Don't work around it or retry blindly.
37
38	## Core workflow
39
40	```
41	1. n8n_instances({mode:"list"}) # see available[] + current + default
42	2. n8n_instances({mode:"switch", name:"prod"}) # bind THIS session to "prod"
43	→ returns { previous, current }; confirm current.name == "prod"
44	3. (do your work) n8n_list_workflows / n8n_get_workflow / n8n_manage_datatable / ...
45	4. Before a credential write or a delete:
46	n8n_instances({mode:"list"}) → re-confirm current, THEN n8n_manage_credentials({action:"create", ...})
47	```
48
49	To move to another instance, just `switch` again. The whole session follows the switch.
50
51	## The `n8n_instances` tool
52
53	Two modes (`mode` is required and enum-validated):
54
55	- `{mode:"list"}` → `{ current, default, available }`, no side effects.
56	- `current` and `default` are each one instance `{ id, name, url, isDefault }` (or `null`).
57	- `available` is every instance, each with an extra `isCurrent` boolean. Match by `name`;
58	never hard-code `id`.
59	- `{mode:"switch", name:"<name>"}` → `{ previous, current }`, and binds this session to the named
60	instance. `name` is case-insensitive.
61
62	### Error envelope (from the `n8n_instances` tool)
63
64	Every error returns `{ error: "<CODE>", message, … }`. The ones you'll actually hit:
65
66	\| Code \| When \| What to do \|
67	\|---\|---\|---\|
68	\| `UNKNOWN_INSTANCE` \| `name` matches no instance \| Pick a name from the `available` list in the error payload and retry. \|
69	\| `NAME_REQUIRED` \| `switch` with no `name` \| Re-call with a `name` (the error lists the valid ones in `available`). \|
70	\| `MULTI_INSTANCE_DISABLED` \| multi-instance mode is off \| Th

People who liked this also viewed

6 picks

Type