Track error codes in responses

Use a response template to extract and aggregate fields from response bodies during a load test. The primary use case: find out what error codes your API is returning under load and how often.

1. Create a response template

Save this as response.json. It mirrors the shape of your API's error response:

{
  "error": {
    "code": "{{ STRING }}"
  }
}

{{ STRING }} tells lmn to extract that field as a string and track the frequency distribution of all distinct values across every response.

2. Run with both templates

lmn run -H https://api.example.com/orders \
  -M post \
  -T ./template.json \
  -S ./response.json

-T is the request template (what to send). -S is the response template (what to track).

3. Read the output

After the run, lmn prints a breakdown under Response:

 Response: error.code ──────────────────────────────────────────────────────────────
  OK                    8741  ████████████████████████████
  INSUFFICIENT_BALANCE   987  ███▌
  NOT_FOUND              201  ▊
  RATE_LIMITED            71  ▎

You can immediately see that 10% of requests are hitting INSUFFICIENT_BALANCE — something that would be invisible from status codes alone if the API returns 200 for business errors.

4. Capture in JSON output

lmn run -f lmn.yaml -S ./response.json --output-file result.json

The response_stats field in the JSON report contains the full distribution. See JSON Output Schema.

Config file equivalent

run:
  host: https://api.example.com/orders
  method: post
  template_path: ./template.json
  response_template_path: ./response.json

execution:
  request_count: 1000
  concurrency: 25

See Template Placeholders for {{ FLOAT }} extraction and nested field tracking.