n8n Validation Expert
Explains n8n validation results and guides iterative fixes for problems that can stop workflows from running.
The skill is primarily advisory and read-only, and its review checklist addresses credentials, injection risks, and confirmation before auto-fixing security or connection issues. However, the main auto-fix examples also show applyFixes:true without a consistent confirmation, least-privilege, rollback, or data-flow procedure. Sensitive-data handling is incomplete, and publisher identity is unverified, so 15 points are deducted.
The documentation is extensive and includes error taxonomies, recovery strategies, abnormal-case guidance, and failure messages, making the main path plausible from static inspection. Execution was not performed, and the material contains inconsistencies: some sections imply expression references are validation-detected while others state they fail only at execution. Version-dependent behavior and tool semantics also lack reproducible evidence, so the static cap requires a conservative deduction.
Trigger phrases, scenarios, validation profiles, and warning-handling guidance are clearly described for Claude Code users operating n8n-mcp. Non-fit boundaries, exact input/output contracts, Chinese-language support, and mainland-China network reachability are not addressed, and several claims depend on n8n-mcp 2.63.0 or later, so 5 points are deducted.
SKILL.md, the error catalog, false-positives guide, and review checklist provide strong progressive disclosure with examples, limitations, related skills, MIT-license context, author information, and a version-history entry. Maintenance ownership, update path, changelog detail, and dependency compatibility are incomplete, while several external statistics and references are not verifiable within the supplied skill evidence, so 6 points are deducted.
The error categories, configuration examples, validation loop, profile guidance, and recovery operations are directly useful for the core troubleshooting task and show meaningful practical value. However, examples and claims were not executed, some advice is server-version-specific or delegated to other skills, and users must still verify behavior in their n8n environment, so 9 points are deducted.
The files provide many code examples and version qualifiers, offering limited static auditability. They contain no committed test suite, CI coverage, third-party execution records, or independent corroboration. Key claims such as 7,841 telemetry occurrences, error percentages, and 94% success within three iterations cannot be verified from the supplied material, warranting only a low score.
- Do not treat the telemetry figures, success rate, error percentages, or claimed real examples as independently verified evidence.
- Auto-fixes can alter workflows; before applying them in production, confirm permissions, scope, credentials, rollback, and user approval.
- Core behavior depends on the n8n-mcp version, especially 2.63.0 or later; verify profile behavior, expression handling, and sanitization on the target server.
- The skill does not establish Chinese-language quality or service reachability from mainland-China networks.
What does this skill do, and when should you use it?
This is one focused skill from the n8n-skills repository, dedicated to validating n8n nodes and workflows. It distinguishes execution-blocking errors from actionable warnings and optional suggestions, and explains the minimal, runtime, ai-friendly, and strict profiles. It also covers validation loops, auto-sanitization, auto-fix behavior, and common workflow-structure problems. It fits users building or reviewing n8n workflows with n8n-mcp in Claude Code, Claude.ai, or another Agent Skills-compatible environment.
Guides users through the valid, errors, warnings, suggestions, and summary fields returned by validate_node and validate_workflow; explains missing_required, invalid_value, type_mismatch, invalid_expression, invalid_reference, and patchNodeField errors; recommends get_node, cleanStaleConnections, and n8n_autofix_workflow; explains profile selection, connection and expression fixes, warning triage, and the use of the referenced workflow-review checklists.
- A user has configured an n8n node with n8n-mcp and needs to understand a missing-field or type-validation error.
- A workflow fails validation because of a broken node reference, invalid expression, or connection problem.
- A user needs to choose between runtime, ai-friendly, and strict validation profiles for development or production hardening.
- Validation returns warnings about error handling, retries, rate limits, or unbounded queries, and the user needs to decide what is worth fixing.
- A workflow update contains issues that n8n_autofix_workflow may resolve, and the user wants to preview or apply those fixes.
- An existing workflow passes validation but still needs review for silent connection bugs, injection risks, or missing error paths.
What are this skill's strengths and limitations?
- Covers node validation, whole-workflow validation, profiles, error categories, and iterative repair.
- Clearly separates execution-blocking errors, security and deprecation warnings, best-practice advisories, and optional optimizations.
- Includes concrete usage patterns for n8n_autofix_workflow, cleanStaleConnections, and save-time auto-sanitization.
- Documents validator behavior and classic false-positive changes for n8n-mcp 2.63.0 and later.
- It is a guidance skill, not an independent n8n validation engine, and requires the n8n-mcp MCP server.
- The source provides no standalone test results or separate installation package for this skill.
- Auto-fix does not cover every problem, including missing-node connections, branch-count mismatches, and some corrupt states.
- Its guidance is centered on n8n-mcp workflow scenarios rather than generic workflow validation.
How do you install this skill?
First install and configure the n8n-mcp MCP server and access to Claude Code, Claude.ai, or the Claude API. To install the repository collection, use /plugin install czlonkowski/n8n-skills; alternatively clone it and run cp -r n8n-skills/skills/* ~/.claude/skills/, then reload Claude Code. To use only this skill, install skills/n8n-validation-expert/ as a skill folder in the target client's supported skills directory; the source does not document a separate copy path for other clients.
How do you use this skill?
In an environment with n8n-mcp configured, ask a relevant question such as “Why is this n8n workflow validation failing?” or “Explain these validate_workflow warnings and tell me which ones require fixes.” The skill activates for validation failures, warnings, profile questions, validation-result interpretation, and auto-fix guidance. The usual process is validate → read errors → fix → validate again, typically for two or three iterations.
How does this skill compare with similar options?
Compared with reading raw validation output alone, this skill adds error classification, profile selection, iterative repair guidance, and auto-fix instructions; the source provides no direct comparison with other dedicated validation products.