n8n MCP Community
Manage self-hosted n8n Community Edition through 44 MCP tools for workflows, executions, credentials, tags, users, and diagnostics, with read-only/write/unsafe safety modes and no external AI calls.
Workflows: list, get, create, update, update one node, delete, activate/deactivate, get version/diff, manage tags, archive/unarchive.
Executions: list, get, delete, retry, stop.
Credentials: schema, list, get, create, update, delete, test, find usage; secret values are never returned.
Tags: list, get, create, rename/update, delete.
Users: list, get, invite, delete.
Diagnostics/discovery: health check, insights summary, security audit, workflow search, bundled node docs, observed node-type inventory, deterministic workflow introspection, community package metadata.
Resources/prompts: 5 static resources (usage guide plus 4 node docs) and 4 prompts for workflow creation, debugging, optimization, and credential management.
Safety: default read-only mode; write/unsafe gates; same-origin bounded requests; redaction/truncation; no inbound HTTP listener or external model calls.
Provides tools for managing n8n Community Edition workflows, nodes, executions, credentials, tags, users, and instance diagnostics through the n8n Public API.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@n8n MCP Community list my workflows"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
n8n MCP Community
A security-focused Model Context Protocol server for self-hosted n8n Community Edition
Give your AI client 44 tools for self-hosted n8n Community Edition. Manage workflows, nodes, runs, credentials, tags, users, and diagnostics. No paid n8n features. The server makes no external AI calls.
Install
You need Node.js 22 or 24, your n8n URL, and an n8n API key.
Install
npm install --global n8n-mcp-community@latest--global installs the command once. Use npx below if you do not want a
global install.
Connect
{
"mcpServers": {
"n8n-community": {
"command": "n8n-mcp-community",
"env": {
"N8N_API_URL": "https://n8n.example.com",
"N8N_API_KEY": "replace-with-a-dedicated-api-key",
"N8N_MCP_MODE": "read-only"
}
}
}
}Restart your client. You should see 44 tools, 5 resources, and 4 prompts.
If the client cannot find the command, run command -v n8n-mcp-community on
macOS/Linux or where n8n-mcp-community on Windows. Put that full path in
"command".
Want no global install? Copy this command into the client configuration shown in the full guide:
npx --yes n8n-mcp-community@latestWant the Claude Desktop app flow? Use the MCPB below. See the full installation guide for updates, rollback, and checks.
Claude Desktop: install the MCPB (no Node.js needed)
Open the latest GitHub release.
Download the file that ends in
.mcpb.In Claude Desktop, open Settings → Extensions → Advanced settings → Install Extension….
Pick the file. Enter your n8n URL, API key, and mode.
Anthropic's MCPB guide explains the same steps. Our installation guide explains how to check the download.
Install public builds only from npm or the latest GitHub release.

Related MCP server: n8n-MCP
Why this project
Community Edition first. The release surface is designed for self-hosted n8n Community Edition, without presenting paid-only capabilities as available.
44 useful tools, one explicit contract. Every tool has bounded inputs, documented side effects, MCP safety annotations, and contract tests.
Safe by default. Read-only mode is the default. You choose when to allow writes or the full tool set.
Deterministic Introspect.
n8n_introspectruns a local 23-rule engine; it does not execute workflows, call an agent, or contact an external model.Surgical node updates.
n8n_update_nodechanges one validated node path while preserving the rest of the workflow and disclosing the Public API's non-atomic limitation.Truthful discovery.
n8n_list_node_typesreports only types observed in accessible workflows. It does not claim to be a complete installed catalog.Data minimization. Credential values, raw execution values, pin data, and static workflow data are not returned by the generic public tools.
Current verification
The current source candidate has been verified with:
exactly 44 tools, 5 resources, and 4 prompts over real stdio;
320 passing tests locally on Node.js 24.21.0; CI and release gates require the pinned Node.js 22.23.1 and 24.18.0 matrix;
zero findings from Gitleaks, Semgrep, Trivy, and both project-root production/full
npm auditruns; all three source scanners are reproduced in CI, with immutable scanner/action identities;a separate no-override consumer install that resolves the patched
@hono/node-server2.x line, requires a zero-finding production audit, and proves the stdio runtime and exact 44/5/4 inventory;a reproducible dependency-license gate covering 187 installed package paths;
bounded same-origin HTTP contracts and zero-request policy-denial tests;
all 44 compiled tool lifecycles on disposable n8n Community 2.30.5 and 2.30.7 instances with egress isolation, revoked keys, cleanup, and zero residue; and
a reviewed offline npm-tarball installation under the project's committed dependency policy, plus a byte-reproducible unsigned MCPB with the same compiled runtime and complete production dependency notices.
Publication proceeds only through the reviewed release procedure: the repository-pinned MCPB signing identity, a human artifact-baseline receipt, tag-scoped protected-environment approval, and an externally signed MCPB handoff verified byte for byte against the reviewed unsigned candidate.
Safety modes
Mode | What it allows | Required configuration |
Read-only | Read-only tools only | Default, or |
Write | Read-only and ordinary mutation tools |
|
Unsafe | All tools, including destructive or external-contact operations |
|
Selecting unsafe mode is the complete server-side authorization gate for unsafe tools. MCP safety annotations remain accurate so compatible clients can apply their native approval UX, and the n8n API key remains the final upstream permission boundary.
Plain HTTP is accepted automatically only for loopback URLs. A non-loopback
HTTP instance additionally requires N8N_ALLOW_INSECURE_HTTP=1, which accepts
the risk of exposing the API key and data in transit.
Tool catalog
Every entry links to its complete input, output, endpoint, compatibility, failure, and security reference.
Workflows
Tool | Mode | Purpose |
Read-only | List accessible workflows with bounded filters and pagination. | |
Read-only | Read one workflow while withholding pin and static data values. | |
Write | Create a validated workflow through the Public API. | |
Write | Guard and update selected fields while preserving omitted fields. | |
Write | Change one validated node property with non-atomic concurrency guards. | |
Unsafe | Permanently delete one workflow. | |
Unsafe | Activate one workflow. | |
Unsafe | Deactivate one workflow. | |
Read-only | Retrieve one retained historical workflow version. | |
Read-only | List tags assigned to a workflow. | |
Write | Replace a workflow's complete tag assignment. | |
Unsafe | Archive one workflow. | |
Unsafe | Restore one archived workflow. | |
Read-only | Compare nodes and connections without returning raw values. |
Executions
Tool | Mode | Purpose |
Read-only | List execution metadata with value-free data-presence summaries. | |
Read-only | Read one execution's metadata without returning workflow payloads. | |
Unsafe | Permanently delete a saved execution. | |
Unsafe | Retry an eligible saved execution. | |
Unsafe | Stop one stoppable execution and bind identity from validated input. |
Credentials
Tool | Mode | Purpose |
Write | Create a credential without returning its values. | |
Unsafe | Permanently delete one credential. | |
Read-only | Read the public schema for a credential type. | |
Read-only | List credential metadata verified on n8n 2.30.5 and 2.30.7. | |
Read-only | Read one credential's public metadata. | |
Write | Update credential metadata or values without echoing secrets. | |
Unsafe | Test a stored credential, potentially contacting its external service. | |
Read-only | Find exact credential-ID references in one bounded workflow page. |
Tags
Tool | Mode | Purpose |
Read-only | List workflow tags. | |
Read-only | Read one tag. | |
Write | Create a tag with the live 1–24 character bound. | |
Write | Rename a tag with the live 1–24 character bound. | |
Unsafe | Permanently delete one tag. |
Users
Tool | Mode | Purpose |
Read-only | List users visible to the API key. | |
Read-only | Read one user by stable ID or exact email. | |
Unsafe | Invite a member or admin. | |
Unsafe | Delete one API-eligible user without unsupported transfer claims. |
Diagnostics and instance metadata
Tool | Mode | Purpose |
Read-only | Perform a bounded same-origin health check. | |
Read-only | Read the official insights summary with optional date filters. | |
Unsafe | Generate n8n's instance security audit. | |
Read-only | Search one workflow page locally by name, node type, or tag. | |
Read-only | Read one of four immutable offline core-node references. | |
Read-only | Inventory node types observed in bounded accessible workflow pages. | |
Read-only | Run deterministic local workflow and execution diagnostics. | |
Read-only | List installed community-package metadata without mutation paths. |
Resources and prompts
The server also exposes five static resources:
n8n://usage-guiden8n://node-docs/webhookn8n://node-docs/coden8n://node-docs/http-requestn8n://node-docs/if
Four prompts guide safe workflow creation, debugging, optimization, and
credential management: create-workflow, debug-workflow,
optimize-workflow, and manage-credentials.
Security model in one minute
The server uses local stdio; it does not expose an inbound HTTP listener.
Connected requests are same-origin and limited to the configured n8n host.
Redirects are rejected; request and response bodies are bounded to 2 MiB.
Generic output is validated, sanitized, bounded to 256 KiB, and wrapped as
{ "data": ..., "redacted": boolean, "untrusted": true }.Raw credential values and generic execution payload values are never returned.
n8n_introspectuses fixed local rules and has no external-model code path.Security controls reduce risk; they do not make untrusted n8n content safe to execute or authorize broader API-key permissions.
Read the complete security model and use SECURITY.md to report a vulnerability privately.
Documentation
Development
npm ci
npm run verify:contributor
npm run sbom > sbom.cdx.jsonnpm run verify:contributor runs the complete keyless contributor gate,
including formatting, dependency licenses/notices, strict TypeScript, all
tests, the compiled runtime, and documentation inventory parity.
See CONTRIBUTING.md before opening a change.
Scope and non-affiliation
This project is not affiliated with, endorsed by, or sponsored by n8n GmbH.
“n8n” is used only to identify compatibility with the n8n product and Public
API. The project does not redistribute the n8n-nodes-base catalog and does not
use browser cookies, interactive session routes, or runtime package downloads to
construct one.
The v0.3.x surface intentionally excludes arbitrary workflow execution, credential/workflow transfer, folders, data tables, beta evaluation endpoints, and execution annotations. See the roadmap for the annotation proposal retained outside the release target.
License and maintainer
The project is released under the MIT License. Third-party dependencies retain their own licenses.
Created and maintained by Dr. Walter Zamarian Jr..
Available Tools
44 toolsn8n_audit_generateGenerate security auditARead-onlyIdempotent
Run one read-only n8n security scan. Omit categories to use n8n's complete default, or provide 1-5 unique risk areas; an empty array is rejected. It makes exactly one POST /audit request and never changes resources or configuration. It is safe to retry but remains unsafe-mode and owner-authorized because the report can expose sensitive security posture. Use n8n_introspect for one workflow. Returns a sanitized, untrusted report; a clean scan returns an empty map. If it cannot fit, retry with fewer categories or read the full report in n8n.
Inputs: categories. Returns: data, redacted, untrusted
n8n_community_packages_listList community packagesARead-onlyIdempotent
List bounded metadata for installed n8n community packages. This call accepts no arguments, filters, cursor, or limit; it keeps at most 100 endpoint records and reports exact omissions. Use n8n_list_node_types for types observed in workflows. It does not download or inspect contents. Requires package-list permission, never installs, updates, or removes packages, and returns untrusted metadata with author emails redacted.
Inputs: no named parameters. Returns: data, redacted, untrusted
n8n_credentials_createCreate credentialA
Create and persist one credential as an additive write. First call n8n_credentials_schema with the planned type, then build data from its field contract; name is only the stored label. isResolvable is forwarded only when supplied. Use n8n_credentials_update for an existing credential. Requires write/unsafe mode plus create permission. Secret values belong only in data input and are excluded from logs and output; returns metadata only.
Inputs: data (required), name (required), type (required), isResolvable. Returns: data, redacted, untrusted
n8n_credentials_deleteDelete credentialADestructive
Permanently delete one stored credential, which can break every referencing workflow. credentialId must be the stable ID from n8n_credentials_list, get, or create; a credential name or type is not accepted. Scan all n8n_credentials_usage pages first, or update when replacement is enough. No secret or rollback is returned. Requires unsafe mode plus credential-delete permission; returns the request-bound credentialId with deleted=true.
Inputs: credentialId (required). Returns: data, redacted, untrusted
n8n_credentials_getGet credentialARead-onlyIdempotent
Get public metadata for one stored credential. credentialId is the stable ID from n8n_credentials_list or create, not its name, type, or secret data. Use n8n_credentials_schema for type fields. The endpoint does not retrieve credential values or test the external service. Requires credential-read permission; returns validated identity, type, flags, and timestamps.
Inputs: credentialId (required). Returns: data, redacted, untrusted
n8n_credentials_listList credentialsARead-onlyIdempotent
List one Public API page of credential metadata; the endpoint is supported from n8n Community 2.30.5. Omit cursor for page one. For later pages, pass nextCursor back unchanged; limit sizes only that request, and the tool never auto-paginates. Use n8n_credentials_get when an ID is known. Requires credential-list permission; returns metadata and nextCursor, never stored values.
Inputs: limit, cursor. Returns: data, redacted, untrusted
n8n_credentials_schemaGet credential schemaARead-onlyIdempotent
Get n8n's Public API field contract for one credential type. credentialType is the public type name used by create/update, not a stored credential ID, name, or secret. Its result tells the caller which keys belong in data; it does not read or test a stored credential. Use n8n_credentials_get for stored metadata. Requires schema-read permission; returns the validated upstream field contract.
Inputs: credentialType (required). Returns: data, redacted, untrusted
n8n_credentials_testTest credentialADestructive
Test one already stored credential by allowing n8n to contact its external service, which receives and may log the attempt. credentialId is the stable ID from n8n_credentials_list or get; inline credential data, names, and types are not accepted. Use get for metadata-only inspection and skip this tool when network contact is unwanted. Requires unsafe mode plus test permission; returns only the target-bound OK/Error outcome and withholds the upstream diagnostic message.
Inputs: credentialId (required). Returns: data, redacted, untrusted
n8n_credentials_updateUpdate credentialADestructive
Update one stored credential, potentially breaking dependent workflows. credentialId selects the record; omit fields to keep them and supply at least one change. If type changes, data is also required and must match n8n_credentials_schema. With data, isPartialData=false replaces the saved field set, while true requests a partial merge. Use create for a new credential. Requires write/unsafe mode plus update permission; returns metadata without secret values.
Inputs: credentialId (required), data, name, type, +3 more. Returns: data, redacted, untrusted
n8n_credentials_usageFind credential usageARead-onlyIdempotent
Scan one workflow page for exact references to a credential. credentialId is the stable ID from n8n_credentials_list or get. Omit cursor for page one; keep credentialId, active, and limit unchanged when passing nextCursor. Unresolved legacy references are counted but never matched, and coverage is complete only when nextCursor is null. Use this before update/delete. Requires workflow-list permission; returns bounded workflow/node matches and omission counts.
Inputs: credentialId (required), limit, active, cursor. Returns: data, redacted, untrusted
n8n_executions_deleteDelete executionADestructive
Permanently delete one saved execution and its retained history. executionId must be the execution record ID from n8n_executions_list or get; a workflow ID is not accepted. Confirm the target with get first. The server provides no recovery or rollback. Requires unsafe mode plus execution-delete permission; returns the request-bound executionId with deleted=true.
Inputs: executionId (required). Returns: data, redacted, untrusted
n8n_executions_getGet executionARead-onlyIdempotent
Get metadata for one saved execution. executionId is the execution record ID from n8n_executions_list, not its workflowId. includeData=true asks n8n whether saved payload data exists, but the MCP still withholds every node input and output; false skips that request detail. Use list for discovery. Requires execution-read permission; returns identity, status, timing, retry fields, and value-free dataPolicy.
Inputs: executionId (required), includeData. Returns: data, redacted, untrusted
n8n_executions_listList executionsARead-onlyIdempotent
List one Public API page of saved execution metadata. Use it for discovery and bounded triage; use n8n_executions_get when an ID is known. status and workflowId filter upstream, cursor resumes one prior page, and includeData only reports whether data exists—values remain withheld. Requires execution-list permission and never auto-paginates; returns metadata and nextCursor.
Inputs: limit, cursor, status, workflowId, +1 more. Returns: data, redacted, untrusted
n8n_executions_retryRetry executionADestructive
Retry one saved execution that has retry data and did not finish successfully; n8n rejects queued/new executions and successful finished executions. executionId comes from n8n_executions_get or list. loadWorkflow=true uses the currently saved workflow, while false uses the original execution snapshot. A retry may repeat earlier external effects; use stop for a running execution. Requires unsafe mode and retry permission; returns metadata without payload values.
Inputs: executionId (required), loadWorkflow. Returns: data, redacted, untrusted
n8n_executions_stopStop executionADestructive
Request cancellation of one stoppable execution: new, unknown, waiting, or running. executionId must be its record ID from n8n_executions_list or get, not a workflow ID. n8n rejects an execution that is already terminal. From a successful response, canceled maps to stopped, another terminal status maps to already_finished, and an unclear body maps to unknown; HTTP 200 alone never proves cancellation. Completed external effects remain. Requires unsafe mode and stop permission; returns the mapped state.
Inputs: executionId (required). Returns: data, redacted, untrusted
n8n_get_node_docsGet node documentationARead-onlyIdempotent
Return one immutable bundled reference. node is an exact local key—webhook, code, http-request, or if—not an arbitrary n8n node-type string. Use n8n_list_node_types to inventory observed types and official docs for anything else. Requires no n8n URL, API key, network, or elevated mode; it never follows the included URL. Returns source/fetched provenance, canonical type, title, summary, guidance, and official URL.
Inputs: node (required). Returns: data, redacted, untrusted
n8n_healthCheck n8n healthARead-onlyIdempotent
Perform one redirect-free same-origin /healthz request with a 10-second timeout. This tool accepts no arguments and requires the configured n8n URL and API key; remote plaintext HTTP also requires N8N_ALLOW_INSECURE_HTTP=1. Use it for HTTP reachability, then use n8n_workflows_list to test API-key capability. No workflow data is read or changed. Returns ok=true and the successful status, never the upstream body.
Inputs: no named parameters. Returns: data, redacted, untrusted
n8n_insights_summaryGet insights summaryARead-onlyIdempotent
Get n8n's official instance insights aggregates. Use it when the Community endpoint exists; use n8n_introspect for one workflow's diagnostics. startDate and endDate are inclusive offset-aware bounds, and supplying both requires startDate <= endDate; omitting both requests n8n's default range. Requires insights-read permission and never fabricates unavailable metrics; returns totals, failures, rates, time saved, and runtime aggregates.
Inputs: endDate, startDate. Returns: data, redacted, untrusted
n8n_introspectInspect n8n workflowARead-onlyIdempotent
Inspect one workflow and saved executions with 23 deterministic local rules. profile sets paired defaults: quick uses 24h/20 and caps maxExecutions at 25; deep uses 168h/50 and reads at most four redacted details. includeSanitizedLabels=false keeps labels opaque; true opts into bounded sanitized labels. Use n8n_audit_generate for instance security or n8n_workflows_get for raw structure. Requires workflow/execution read permission; never executes or calls AI; returns findings and coverage.
Inputs: workflowId (required), profile, lookbackHours, maxExecutions, +1 more. Returns: sample, status, metrics, profile, +8 more
n8n_list_node_typesList observed node typesARead-onlyIdempotent
Inventory node types observed across workflow pages; this proves usage, not installation. Omit cursor to start at page one. maxPages limits how many pages this call follows; active and limit apply to every page, so keep them unchanged when continuing from nextCursor. Complete coverage requires starting at page one and reaching the end. Use n8n_get_node_docs for bundled references. Requires workflow-list permission; returns counts and exact coverage, capped at 30 seconds or 20,000 nodes.
Inputs: limit, active, cursor, maxPages. Returns: data, redacted, untrusted
n8n_search_workflowsSearch workflowsARead-onlyIdempotent
Search one workflow page for a case-insensitive substring. query is matched locally only in the fields named by searchIn; active filters upstream first. Omit cursor for page one, then keep query, searchIn, active, and limit unchanged when passing nextCursor. This is not a global index; use n8n_workflows_list for an unsearched page. Requires workflow-list permission; returns at most 50 value-free matches and explicit scan state.
Inputs: query (required), limit, active, cursor, +1 more. Returns: data, redacted, untrusted
n8n_tags_createCreate tagA
Create and persist one reusable workflow label as an additive write. name is the literal display label, not a tag ID or workflow assignment; it must already be trimmed. Duplicate-name handling belongs to n8n, and creation never assigns the tag to a workflow. Use n8n_tags_update for an existing tag and n8n_workflows_update_tags to assign it. Requires write/unsafe mode plus create permission; returns validated metadata.
Inputs: name (required). Returns: data, redacted, untrusted
n8n_tags_deleteDelete tagADestructive
Permanently delete one workflow tag, which can remove that label from many workflows. tagId must be the stable ID from n8n_tags_list or get; a tag name or workflow ID is not accepted. Review the target and affected workflows first, or use n8n_tags_update when a rename is enough. No rollback is provided. Requires unsafe mode plus tag-delete permission; returns the request-bound tagId with deleted=true.
Inputs: tagId (required). Returns: data, redacted, untrusted
n8n_tags_getGet tagARead-onlyIdempotent
Get one workflow tag. tagId is the stable tag identity returned by n8n_tags_list or create; a tag name or workflow ID is not accepted. Use it before assigning, renaming, or deleting the tag. The ID identifies the reusable tag itself, not an assignment, so this call neither lists nor changes assigned workflows. Requires tag-read permission; returns validated ID, name, and optional timestamps.
Inputs: tagId (required). Returns: data, redacted, untrusted
n8n_tags_listList tagsARead-onlyIdempotent
List one Public API page of workflow tags. Omit cursor for page one; for later pages, pass nextCursor back unchanged. limit sizes only the current page, so continue until nextCursor is null for complete coverage. Use n8n_tags_get when the stable tag ID is known. Requires tag-list permission and never changes assignments; returns validated metadata and nextCursor.
Inputs: limit, cursor. Returns: data, redacted, untrusted
n8n_tags_updateUpdate tagADestructive
Rename one existing workflow tag while preserving its ID and assignments. tagId is the stable selector from n8n_tags_list or get; name is the complete new display label, not another ID. Duplicate-name acceptance belongs to n8n. Use n8n_tags_create for a new label and n8n_workflows_update_tags to change assignments. Requires write/unsafe mode plus tag-update permission; returns validated metadata.
Inputs: name (required), tagId (required). Returns: data, redacted, untrusted
n8n_update_nodeUpdate one workflow nodeADestructive
Update one path on one node while preserving sibling data. workflowId selects the current workflow, nodeId selects exactly one node inside it, and path selects the mutable field whose contract validates value. Copy expectedVersionId from the latest workflow read and set acknowledgeNonAtomicRisk=true; both guards are required before the full PUT. Use n8n_workflows_update for top-level or multi-node edits. Requires write/unsafe mode plus read and update permission; returns the changed path and residual race risk.
Inputs: path (required), value (required), nodeId (required), workflowId (required), +2 more. Returns: data, redacted, untrusted
n8n_users_createInvite userA
Create a pending invitation as an additive write that may send email. email is the future account login; use n8n_users_get first when it may exist. role is a global account role, not project membership, and omission selects global:member; owner invitations are rejected. Requires unsafe mode plus invite permission. An inconclusive response may still mean a pending user exists; returns delivery state but never the acceptance URL.
Inputs: email (required), role. Returns: data, redacted, untrusted
n8n_users_deleteDelete userADestructive
Permanently delete one API-eligible user. userId must be the stable ID from n8n_users_list or get; an email address is not accepted. The input has no transfer target, so workflow or credential ownership handling remains entirely with n8n. Verify the target with get first. Requires unsafe mode plus user-delete permission; returns the request-bound userId with deleted=true and provides no rollback claim.
Inputs: userId (required). Returns: data, redacted, untrusted
n8n_users_getGet userARead-onlyIdempotent
Get one user by stable ID or exact email. userIdOrEmail routes ID-shaped input to ID lookup and other input to an exact percent-encoded email lookup; partial email matching is not used. includeRole=true asks n8n for the global role, but permissions may still hide it. Use n8n_users_list for discovery. Requires user-read permission; returns redaction-protected identity and account state without mutation.
Inputs: userIdOrEmail (required), includeRole. Returns: data, redacted, untrusted
n8n_users_listList usersARead-onlyIdempotent
List one Public API page of users. Use it for discovery; use n8n_users_get when a stable ID or exact email is known. includeRole=true only asks n8n for roles, which may still be omitted by permission; cursor resumes one prior page and this call never auto-paginates. Requires user-list permission; returns redaction-protected metadata and nextCursor.
Inputs: limit, cursor, includeRole. Returns: data, redacted, untrusted
n8n_workflows_activateActivate workflowADestructive
Activate one workflow so production triggers can accept future events; this does not execute it now. workflowId must be the current stable ID from n8n_workflows_list or get, not a name or version ID. Review the workflow first; use n8n_workflows_deactivate to stop future triggers. Requires unsafe mode plus activation permission; returns target-validated active state metadata.
Inputs: workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_archiveArchive workflowADestructive
Archive one workflow without deleting its saved definition. workflowId must be the current stable ID from n8n_workflows_list or get, not a name or version ID. Use this instead of delete for reversible removal; use deactivate only to stop triggers. The availability change can disrupt callers. Requires unsafe mode plus archive permission; returns target-validated metadata with isArchived=true.
Inputs: workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_createCreate workflowA
Create and persist one workflow as an additive write without activating its triggers. name, nodes, and connections define a new record; nodes and connections must form one complete graph, while omitting settings uses an empty object. Optional pinData, staticData, description, and nodeGroups are forwarded only when supplied. Use n8n_workflows_update for an existing workflow and activate for triggers. Requires write or unsafe mode plus create permission; returns the validated projection without pin/static values.
Inputs: name (required), nodes (required), connections (required), pinData, +4 more. Returns: data, redacted, untrusted
n8n_workflows_deactivateDeactivate workflowADestructive
Deactivate one workflow's future production triggers without deleting saved data or stopping runs already in progress. workflowId must be the current stable ID from n8n_workflows_list or get, not a name or version ID. Use n8n_workflows_activate to reverse the state or archive for lifecycle removal. Requires unsafe mode plus deactivation permission; returns target-validated inactive state metadata.
Inputs: workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_deleteDelete workflowADestructive
Permanently delete one workflow and its saved definition. workflowId must be the current stable ID from n8n_workflows_list or get; a name or version ID is not accepted. Use n8n_workflows_archive for reversible removal and n8n_workflows_get for inspection. This action has no rollback or transfer. Requires unsafe mode plus delete permission; returns the request-bound workflowId with deleted=true.
Inputs: workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_diffCompare workflow versionsARead-onlyIdempotent
Compare two states of one workflow without returning changed values. workflowId selects the workflow; fromVersionId is the retained baseline, and toVersionId is a different retained target. Omit toVersionId to compare with current state. ignoreLayout=true hides position-only changes. Use n8n_workflows_get_version for raw history. Requires workflow/history read permission; returns coverage, counts, and a byte-bounded prefix of up to 200 changes.
Inputs: workflowId (required), fromVersionId (required), toVersionId, ignoreLayout. Returns: data, redacted, untrusted
n8n_workflows_getGet workflowARead-onlyIdempotent
Get one current workflow for inspection or reviewed editing. workflowId is the stable ID returned by n8n_workflows_list or create, not a workflow name or version ID. Use n8n_workflows_get_version for history. excludePinnedData changes only the upstream request and the pinDataPresent/staticDataPresent flags; pinned values are withheld whether true or false. Requires permission to read the target; returns structure, state, settings, and presence flags.
Inputs: workflowId (required), excludePinnedData. Returns: data, redacted, untrusted
n8n_workflows_get_tagsGet workflow tagsARead-onlyIdempotent
List tags currently assigned to one workflow. workflowId selects the workflow from n8n_workflows_list or get; it is not a tag ID and this call does not list unassigned tags. Use it before n8n_workflows_update_tags when assignments must be preserved; use n8n_tags_list for all tags. The endpoint has no cursor, so the tool keeps at most 100 records and reports exact omissions. Requires assignment-read permission; returns validated tag metadata.
Inputs: workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_get_versionGet workflow versionARead-onlyIdempotent
Get one retained workflow snapshot. workflowId is the stable workflow ID; versionId is that workflow's separate retained-version ID. Do not swap them or pass a workflow name. Use n8n_workflows_get for current state or n8n_workflows_diff for comparison. Both returned IDs must match the selectors; a missing endpoint and a pruned version share an explicit ambiguous 404 outcome. Requires history-read permission; returns validated nodes and connections subject to retention.
Inputs: versionId (required), workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_listList workflowsARead-onlyIdempotent
List one Public API page of workflows visible to the configured API key. Use n8n_search_workflows for substring matching or n8n_workflows_get for a known ID. Omit cursor for page one. active, tags, and name are combined upstream; keep those filters and excludePinnedData unchanged when passing nextCursor. limit sizes only that page, and the tool never auto-paginates. Requires workflow-list permission; returns projected workflows and nextCursor, never pin/static values.
Inputs: name, tags, limit, active, +2 more. Returns: data, redacted, untrusted
n8n_workflows_unarchiveUnarchive workflowADestructive
Restore one archived workflow without activating its triggers. workflowId must identify the archived workflow itself, not a name or version ID. Use this to reverse n8n_workflows_archive; call activate separately only if triggers should resume. Current active state is not inferred. Requires unsafe mode plus unarchive permission; returns target-validated metadata with isArchived=false.
Inputs: workflowId (required). Returns: data, redacted, untrusted
n8n_workflows_updateUpdate workflowADestructive
Update selected top-level fields through a non-atomic full PUT. workflowId selects the current record; copy expectedVersionId from its latest read, and supply at least one writable field. The version must match both pre-write reads. Omitted fields are preserved, but each supplied array or object replaces that whole field. Use n8n_update_node for one node property and n8n_workflows_update_tags for tags. Requires write/unsafe mode plus read and update permission; returns the confirmed projection.
Inputs: workflowId (required), expectedVersionId (required), name, nodes, +6 more. Returns: data, redacted, untrusted
n8n_workflows_update_tagsUpdate workflow tagsADestructive
Replace one workflow's complete tag assignment; this is not a merge. workflowId selects the workflow, while every tagIds item must be a stable tag ID from n8n_tags_list or get; tag names are not accepted. tagIds is the whole desired set, so omitted IDs are removed and [] clears all tags. Use n8n_workflows_get_tags first to preserve assignments. Requires write/unsafe mode plus assignment permission; returns n8n's validated replacement.
Inputs: tagIds (required), workflowId (required). Returns: data, redacted, untrusted
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 44 tools
Each tool targets a distinct resource/action combination, and the descriptions explicitly disambiguate near neighbors such as list vs get vs search, get_version vs diff, archive vs deactivate, and workflows_update vs update_node. No two tools appear to do the same thing.
The dominant n8n_<resource>_<action> pattern is consistent across most tools, e.g. n8n_workflows_get, n8n_tags_delete, n8n_executions_list. A few verb-first outliers like n8n_search_workflows, n8n_update_node, n8n_get_node_docs, and n8n_list_node_types slightly break predictability.
44 tools is a heavy surface for an agent to navigate and exceeds the 25+ threshold, even though the resource coverage is broad. The tools are well-organized, but the sheer count adds meaningful selection overhead.
The server covers full lifecycle operations for workflows, executions, credentials, tags, and users, plus search, audit, insights, and node documentation. Minor gaps exist—such as no workflow execution trigger, no user update, and no package install/removal—but they are workable.
Maintenance
Related MCP Connectors
Security scanner for n8n workflows + live MCP Trust-Check. 18 rules, OWASP mapped. Paid x402 API.
n8n MCP — query your own n8n instance (BYO).
Open-source Zapier/n8n alternative as an MCP server: agents build, run and debug your workflows.
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
Related MCP Servers
- AlicenseBqualityDmaintenanceAn MCP server enabling secure interaction with n8n workflows, executions, and settings via the Model Context Protocol, designed for integration with Large Language Models (LLMs).3348 npm118MIT
- AlicenseAqualityAmaintenanceA Model Context Protocol server that provides AI assistants with comprehensive access to n8n node documentation, properties, and operations for effective workflow automation.2947,555 npm23,286MIT
- AlicenseBqualityCmaintenanceSafe MCP server and CLI tool for testing, scoring, and inspecting n8n workflows. Features 19 tools for workflow testing, execution traces, tiered scoring, and a built-in node catalog with 800+ node types. Designed with safety-first approach: no credential management, no secrets lifecycle, no auto-fix loops.191MIT
- AlicenseBqualityDmaintenanceA comprehensive MCP server that provides full control over n8n automation workflows through natural language. It offers 43 tools for managing workflows, executions, credentials, and data tables, with safety features like write-mode protection and double-validated workflow creation.431MIT