n8n Execution Logs

Fetch recent execution logs from the n8n instance to debug workflows and monitor execution status.

When to Use This Skill (vs MCP)

Two ways to interact with n8n programmatically:

Task	Use
Find a workflow	MCP: `search_workflows`
View workflow nodes	MCP: `get_workflow_details`
Run a workflow	MCP: `execute_workflow`
Debug an execution	This skill or API: `GET /executions/{id}?includeData=true`
List recent failures	This skill with `status=error`
Retry a failed run	API: `POST /executions/{id}/retry`

MCP tools have built-in authentication and work for workflow-level operations. REST API (this skill) is required for execution data - MCP cannot fetch execution logs or outputs.

Configuration

•Environment file: .claude/n8n-api.env
•Variables: N8N_API_URL, N8N_API_KEY

Instructions

When this skill is invoked:

•
Load environment variables from .claude/n8n-api.env:
bash
```
source .claude/n8n-api.env
```

•

Fetch executions using curl:

bash

curl -s "$N8N_API_URL/api/v1/executions?limit=<LIMIT>" \
  -H "X-N8N-API-KEY: $N8N_API_KEY"

Default limit is 3 unless the user specifies otherwise.

•
Parse and display the results in a readable format:

For each execution, show:
- •Execution ID
- •Workflow name (fetch from workflow details if needed)
- •Status (success/error/running/waiting)
- •Started at (human-readable timestamp)
- •Finished at (if completed)
- •Duration (calculated)
- •Error message (if failed)
- •Mode (manual/trigger/webhook)

•

Format output as a table or structured list:

code

## Recent Executions

| # | Workflow | Status | Started | Duration | Mode |
|---|----------|--------|---------|----------|------|
| 1 | workflow-name | Success | 2h ago | 1.2s | trigger |
| 2 | workflow-name | Error | 5h ago | 0.3s | manual |

For failed executions, include the error details below the table.

Getting Execution Details by ID

To investigate a specific execution:

Step 1: Identify the execution ID

•In n8n UI: Executions sidebar shows IDs
•From workflow logs: Execution ID is logged on each run
•From error handler: Captures execution_id of failed runs

Step 2: Fetch with full node data

bash

curl -s "$N8N_API_URL/api/v1/executions/<ID>?includeData=true" \
  -H "X-N8N-API-KEY: $N8N_API_KEY"

Step 3: Understanding the response

Basic fields:

json

{
  "id": "36006",
  "finished": true,
  "mode": "manual",
  "status": "success",
  "workflowId": "S0hk9fjhB-itqaFkq9qnI",
  "startedAt": "2026-01-21T09:56:18.856Z",
  "stoppedAt": "2026-01-21T09:57:39.785Z"
}

With ?includeData=true, the response includes data.resultData.runData containing each node's input/output.

API Endpoints Reference

Endpoint	Method	Purpose
`/api/v1/executions`	GET	List executions (with filters)
`/api/v1/executions/{id}`	GET	Get specific execution details
`/api/v1/executions/{id}/retry`	POST	Retry a failed execution
`/api/v1/workflows`	GET	List all workflows
`/api/v1/workflows/{id}`	GET	Get workflow details
`/api/v1/workflows/{id}/activate`	POST	Activate a workflow
`/api/v1/workflows/{id}/deactivate`	POST	Deactivate a workflow

Arguments

•limit=N - Number of executions to fetch (default: 3, max: 100)
•workflow=ID - Filter by specific workflow ID
•status=success|error|waiting - Filter by status

Example Usage

code

/n8n-executions              # Last 3 executions
/n8n-executions limit=10     # Last 10 executions
/n8n-executions status=error # Recent failures only

Error Handling

•If API returns 401: API key may be invalid or expired
•If API returns 404: Check the n8n instance URL
•If connection fails: Instance may be sleeping (Railways free tier) - retry in 30s