Report formats reference

Every reporter renders the same three events — run started, test finished, run finished — plus test started where the format has a use for it.

Format	Streams	Emitted
human	yes	Per test, plus a digest at the end
jsonl	yes	One object per test, then a summary object
json	no	One document in run_finished
teamcity	yes	Service messages per event
junit	no	One XML document in run_finished

Buffering matters when a run is interrupted: json and junit produce nothing if the process is killed mid-run, while jsonl and teamcity have already flushed what completed.

Statuses

Value	Meaning	Green?
passed	Script finished without throwing	yes
failed	Assertion failed, script threw, or the script file was missing	no
timed_out	Client or interpreter deadline reached	no
skipped	Dry run, or drained after a fail-fast trip	yes

Human

running 3 tests across 1 suite
PASS example.com answers 200 [212ms]
FAIL cart step 1 [28ms]
TIME report generation finishes [10.01s]
SKIP cart step 2 [0ms]

failures:

  cart step 1 (./suite.toml::t1)
    Runtime error: cart service unavailable (line 1, position 1)
    | attempting cart creation

test result: FAILED. 0 passed; 1 failed; 1 timed out; 1 skipped; finished in 10.02s

Element	Rule
Header	running N test(s) across M suite(s); suites are counted by distinct suite path. Suppressed by -q
Mark	PASS (green), FAIL (red), TIME (red), SKIP (yellow)
Duration	<ms>ms under one second, otherwise <s.ss>s
Retries	(after N attempts) when attempts > 1
Captured output	Non-failing tests (passed and skipped) only with -v, prefixed |
Digest	One block per failure: name, qualified id, message lines, captured output
Summary	Always printed; ok when green, FAILED otherwise

With -q, only failing tests print a line — the header and passing lines are suppressed, the digest and summary are not.

Color follows --color: auto (TTY detection), always, or never.

JSONL

One JSON object per line, flushed as each test finishes, then a summary object. Keys are alphabetical.

Test object

{"attempts":1,"duration_ms":7,"id":"demo/suite.toml::assertions","message":null,"name":"assertion helpers, no network","output":["running against local"],"script":"demo/assertions_test.snag","status":"passed","suite":"Snag Demo Regression Suite","suite_path":"demo/suite.toml","tags":["fast"],"test_id":"assertions"}

Field	Type	Description
id	string	Qualified id, <suite_path>::<test_id>
test_id	string	The manifest id
name	string	Display name
suite	string	Suite title
suite_path	string	Manifest path as discovered
script	string	Resolved script path
tags	array of strings	Manifest tags
status	string	passed, failed, timed_out, skipped
duration_ms	integer	Total duration, including every retry attempt
attempts	integer	Attempts made. 0 for skipped tests
message	string or null	Failure message, timeout message, or skip reason
output	array of strings	Captured print / debug / print_response lines

Summary object

{"duration_ms":8,"failed":0,"passed":1,"skipped":0,"timed_out":0,"total":1,"type":"summary"}

Distinguished by "type": "summary" — the field exists only here, so select(.type != "summary") separates the two shapes.

Field	Type	Description
type	string	Always summary
total	integer	Sum of the four counters
passed / failed / timed_out / skipped	integer	Per-status counts
duration_ms	integer	Wall-clock duration of the whole run

JSON

One document, written after the run finishes.

{
  "schema": "snag.run/v1",
  "summary": {
    "duration_ms": 7,
    "failed": 0,
    "passed": 1,
    "skipped": 0,
    "timed_out": 0,
    "total": 1,
    "type": "summary"
  },
  "tests": [
    {
      "attempts": 1,
      "duration_ms": 7,
      "id": "demo/suite.toml::assertions",
      "message": null,
      "name": "assertion helpers, no network",
      "output": ["running against local"],
      "script": "demo/assertions_test.snag",
      "status": "passed",
      "suite": "Snag Demo Regression Suite",
      "suite_path": "demo/suite.toml",
      "tags": ["fast"],
      "test_id": "assertions"
    }
  ]
}

Field	Description
schema	Format version, currently snag.run/v1. Branch on this rather than sniffing the shape
summary	The same object the JSONL summary line carries
tests	Array of test objects, in completion order

Output is pretty-printed. Keys are alphabetical within each object — stable, but generated rather than authored, so parse by key.

TeamCity

Service messages, flushed per event. Understood by TeamCity and by IntelliJ IDEA, which renders a native test tree.

##teamcity[testSuiteStarted name='Snag Demo Regression Suite']
##teamcity[testStarted name='assertion helpers, no network' locationHint='file://demo/assertions_test.snag']
##teamcity[testStdOut name='assertion helpers, no network' out='running against local']
##teamcity[testFinished name='assertion helpers, no network' duration='5']
##teamcity[testSuiteFinished name='snag']

Event	Message
Run started	testSuiteStarted with the first test's suite title, or snag when empty
Test started	testStarted with locationHint='file://<script>'
Captured output	One testStdOut per line, before the status message
Failure / timeout	testFailed with the message
Skip	testIgnored with the message
Test finished	testFinished with duration in milliseconds
Run finished	testSuiteFinished name='snag'

Values are escaped per the TeamCity spec: ' → \|', newline → \|n, carriage return → \|r, \| → \|\|, [ → \|[, ] → \|].

Note that the closing testSuiteFinished always uses the literal name snag, while the opening message uses the suite title.

JUnit

One XML document, written after the run finishes.

<?xml version="1.0" encoding="UTF-8"?>
<testsuites name="snag" tests="1" failures="0" skipped="0" time="0.005">
  <testsuite name="snag" tests="1" failures="0" skipped="0" time="0.005">
    <testcase classname="Snag Demo Regression Suite" name="assertion helpers, no network" time="0.005">
      <system-out>running against local</system-out>
    </testcase>
  </testsuite>
</testsuites>

Element / attribute	Value
testsuites / testsuite name	Always snag; all tests go into one testsuite regardless of how many manifests ran
tests	summary.total()
failures	failed + timed_out — JUnit has no separate timeout concept
skipped	summary.skipped
time	Seconds, three decimals
testcase classname	The suite title
testcase name	The test name
testcase time	Test duration in seconds
<failure message="…" type="…">	For failed and timed_out; type carries which
<skipped message="…"/>	For skipped tests
<system-out>	Captured output, newline-joined, present only when non-empty

Escaping: &, <, >, ", and ' become entities, and control characters that are illegal in XML 1.0 (other than tab, newline, carriage return) are replaced with a space — a binary-ish response body cannot produce an unparseable report.

list output

snag list emits its own shape — a description of tests, not results.

$ snag list demo/suite.toml -t fast -f jsonl
{"id":"demo/suite.toml::assertions","name":"assertion helpers, no network","parallel_safe":true,"script":"demo/assertions_test.snag","suite":"Snag Demo Regression Suite","suite_path":"demo/suite.toml","tags":["fast"],"test_id":"assertions","timeout_ms":10000}

Field	Description
id, test_id, name, suite, suite_path, script, tags	As in a test object
timeout_ms	Resolved timeout in milliseconds, null when none applies
parallel_safe	Effective value after defaults

-f jsonl prints one object per line; json, junit, and teamcity all print a pretty-printed JSON array — list has no XML or service-message rendering. human prints names, or the indented block with --long.

Stability

• Field names and the snag.run/v1 schema string are the contract; key order is alphabetical but not part of it.
• New fields may be added within v1; consumers should ignore unknown keys.
• The human format is for people. Parse jsonl or json instead.

See also

• Reporters guide — choosing and combining formats
• Adding a reporter