reference.md

Reference (types)

type: report files are generated output — not hand-authored. This page documents the MMT YAML schema written by testlight and the VS Code export.

Common fields

• type: report
• kind: functional | load
• name: string — source file or run title
• overview: object — run summary (shape depends on kind)
• cancelled: boolean (optional) — set when the run was stopped early

Functional report (kind: functional)

Used for test and suite runs. Top-level checks lists suite and test entries.

• overview:
  • started_at: ISO timestamp
  • finished_at: ISO timestamp (optional)
  • duration: string (e.g. 1.234s)
  • checks: number — total check/assert steps
  • passed: number
  • failed: number
  • errors: number (always 0 in current output)
  • skipped: number (always 0 in current output)
• checks: array of check entries

Check entry

Each item in checks:

• name: string
• type: suite | test
• file: string (optional) — source .mmt path
• duration: string (optional)
• result: passed | failed
• checks: array (optional) — nested step entries; omitted for type: suite

Suite entries are status-only rows and are not expandable in the report panel or exported HTML.

Step entry (nested under a test)

• name: string — check or assert title
• type: check
• step: assert | debug (optional; omitted for check steps)
• duration: string (optional)
• result: passed | failed
• expects: array (optional) — per-comparison results
  • comparison: string
  • result: passed | failed
  • actual: any (on failure)
  • expected: any (on failure)
• failure: object (optional, on failed steps)
  • message: string
  • actual: string
  • expected: string
  • operator: string

type: report
kind: functional
name: suite.mmt
overview:
  started_at: "2026-03-06T10:30:00.000Z"
  finished_at: "2026-03-06T10:30:01.234Z"
  duration: 1.234s
  checks: 4
  passed: 3
  failed: 1
  errors: 0
  skipped: 0
checks:
  - name: nested-suite.mmt
    type: suite
    result: passed
  - name: test-file.mmt
    type: test
    file: test-file.mmt
    duration: 0.500s
    result: failed
    checks:
      - name: status == 200
        type: check
        result: passed
        duration: 0.100s
      - name: result.name == John
        type: check
        result: failed
        duration: 0.050s
        failure:
          message: "result.name ==: expected John got Jane"
          actual: Jane
          expected: John
          operator: "result.name =="

Load report (kind: load)

Used for load test runs. Load metrics are stored at the report root — not under a nested load element — and load reports do not include per-iteration checks.

• overview: merges timing with load summary fields:
  • started_at: / finished_at: / duration: (same as functional)
  • iterations: number
  • requests: number
  • successes: / failures: number
  • success_rate: / failed_rate: / error_rate: number (0–1)
  • throughput: number (requests per second)
  • data_received: / data_sent: number (optional, bytes)
  • errors: / skipped: number (always 0 in current output)
• test: string — path to the underlying type: test file
• config: object
  • threads: number
  • repeat: string | number
  • rampup: string
• latency: object (milliseconds)
  • min: / avg: / med: / max: / p90: / p95: / p99: number
• http: object (optional)
  • status_codes: record<string, number>
  • failed_requests: number
  • connect_avg: / send_avg: / waiting_avg: / receive_avg: number (optional)
• thresholds: array (optional)
  • name: string
  • expression: string (optional)
  • actual: number (optional)
  • result: passed | failed
• errors: array (optional) — aggregated error messages
  • message: string
  • count: number
  • rate: number (optional)
• snapshots: array — per-second samples
  • at: number — elapsed seconds from run start, starting at 0
  • active_threads: / requests: / errors: / error_delta: / throughput: / response_time: / error_rate: / p95: number (optional)

type: report
kind: load
name: Login Load Test
overview:
  started_at: "2026-05-04T10:30:00.000Z"
  finished_at: "2026-05-04T10:31:00.000Z"
  duration: 1m
  iterations: 1000
  requests: 3000
  successes: 2995
  failures: 5
  success_rate: 0.998
  failed_rate: 0.002
  error_rate: 0.002
  throughput: 50.0
  errors: 0
  skipped: 0
test: ./tests/login.mmt
config:
  threads: 100
  repeat: 1m
  rampup: 10s
latency:
  min: 12
  avg: 48.2
  med: 41
  max: 880
  p90: 92
  p95: 120
  p99: 310
http:
  status_codes:
    "200": 2995
    "500": 5
  failed_requests: 5
thresholds:
  - name: p95 latency
    expression: p95 < 200
    actual: 120
    result: passed
errors:
  - message: HTTP 500
    count: 5
    rate: 0.005
snapshots:
  - at: 0
    active_threads: 20
    requests: 500
    errors: 0
    throughput: 50
    response_time: 48.2
    p95: 110

Exported formats

The MMT schema above is the canonical source. Other formats are projections:

Format	Notes
HTML	Self-contained page with overview cards, suite/test rows, load charts
Markdown	Compact ## Tests section; load reports add Mermaid charts and snapshot tables
JUnit	Standard <testsuites>/<testcase> output; load metrics as <property> values such as load.threads, load.throughput, load.latency.p95, load.snapshots.0.at

See also: Overview · CLI · Suite exports · Load test exports