Prompt file imported from raeef1001/n8n-copilot-skills (.github/prompts/n8n-validation.prompt.md). Copyright stays with the author.

n8n Validation Expert

Validation Philosophy

Validate early, validate often. Expect 2-3 validate → fix cycles (23s thinking, 58s fixing).

Validation Profiles

validate_node({nodeType: "nodes-base.slack", config: {...}, mode: "full", profile: "runtime"})

Error Severity Levels

Errors (Must Fix) - Blocks execution

• missing_required - Required field not provided
• invalid_value - Value not in allowed options
• type_mismatch - Wrong data type
• invalid_reference - Referenced node doesn't exist
• invalid_expression - Expression syntax error

Warnings (Should Fix) - Works but may have issues

• best_practice - Recommended configuration
• deprecated - Using old API/feature
• performance - Potential performance issue

Suggestions (Optional) - Nice to have

• optimization - Could be more efficient
• alternative - Better approach available

The Validation Loop

1. Configure node
   ↓
2. validate_node (mode: "minimal") - quick check
   ↓
3. Read errors, fix config
   ↓
4. validate_node (mode: "full", profile: "runtime")
   ↓
5. Fix remaining errors
   ↓
6. Repeat until valid (2-3 iterations typical)
   ↓
7. validate_workflow - complete workflow check

Common Error Fixes

missing_required

// Error: "Channel name is required"
// Fix: Add the missing field
config.channel = "#general";

invalid_value

// Error: "Operation must be one of: post, update, delete"
config.operation = "post";  // Use valid value

type_mismatch

// Error: "Expected number, got string"
config.limit = 100;  // Number, not "100"

invalid_expression

// Error: "Invalid expression: $json.name"
config.text = "={{$json.name}}";  // Add {{}}

invalid_reference

// Error: "Node 'HTTP Requets' does not exist"
config.expression = "={{$node['HTTP Request'].json.data}}";  // Fix typo

Auto-Sanitization System

Runs automatically on ANY workflow update:

What it fixes:

• Binary operators (equals, contains) → removes singleValue
• Unary operators (isEmpty, isNotEmpty) → adds singleValue: true
• IF/Switch metadata → adds conditions.options

What it CANNOT fix:

• Broken connections → Use cleanStaleConnections operation
• Branch count mismatches → Add missing connections
• Paradoxical corrupt states → Manual intervention

Common False Positives (acceptable to ignore)

1. "Missing error handling" - OK for simple/test workflows
2. "No retry logic" - OK for idempotent operations
3. "Missing rate limiting" - OK for internal APIs
4. "Unbounded query" - OK for small known datasets

Use ai-friendly profile to reduce false positives.

Recovery Strategies

Start Fresh

When config is severely broken: note required fields from get_node, build minimal config, add incrementally.

Clean Stale Connections

n8n_update_partial_workflow({id, operations: [{type: "cleanStaleConnections"}]})

Auto-fix

n8n_autofix_workflow({id, applyFixes: false})  // Preview
n8n_autofix_workflow({id, applyFixes: true})   // Apply

Best Practices

Do:

• Validate after every significant change
• Read error messages completely (they contain fix guidance)
• Fix errors iteratively (one at a time)
• Use runtime profile for pre-deployment
• Trust auto-sanitization for operator issues

Don't:

• Skip validation before activation
• Use strict during development (too noisy)
• Manually fix auto-sanitization issues
• Deploy with unresolved errors