Reading Results

Inspect individual tests, step-by-step execution logs, run history, and failure bundles from the CLI.

Inspecting a test

testsprite test get test_3a9f21c7

id:            test_3a9f21c7
projectId:     proj_8f0f6
name:          Checkout flow — guest user
type:          frontend
createdFrom:   cli
status:        passed
planSteps:     3
  1. [action] Navigate to /checkout as a guest.
  2. [action] Enter the shipping address and click 'Continue'.
  3. [assertion] Verify the order summary shows the correct total.
createdAt:     2026-06-10T08:14:22Z
updatedAt:     2026-06-15T10:04:38Z

List tests in a project, optionally filtered:

testsprite test list --project proj_8f0f6 --type frontend --status failed,blocked

Flag	Values	Description
--type	frontend | backend	Test type
--created-from	portal | mcp | cli	Authoring surface
--status	comma-separated	draft, ready, queued, running, passed, failed, blocked, cancelled, unknown
--env <name>	environment name	Show statuses for that environment (adds an ENV column)
--page-size <n>	1–100, default 25	Items per page
--starting-token <token>	opaque cursor	Continue from a previous page

Status is per environment

test list and test get without --env show each frontend test’s latest result in the project’s default environment, and --status filters on that, not on the latest run in any environment. Use --env <name> to read another environment. A --target-url run is recorded under its temporary environment and does not change the listed status. project list and project get show the default environment and environment count.

Step-by-step results

test steps returns the step log of the latest run, or of a specific run:

testsprite test steps test_3a9f21c7 --run-id run_5c1d9a2b

Each entry has the step name, its status (passed or failed), and its run. Use --page-size and --max-items for long logs. Steps logged before run IDs were tracked are excluded when you pass --run-id.

The latest result

testsprite test result test_3a9f21c7

Without --env or --history, this shows the newest run across all environments, even if still running. In text output, if that run is not on the Default environment, a line follows with the Default environment’s status and the command to read that environment’s history (omitted with --output json, --env, --history, or when the Default status cannot be read). --env requires --history.

--include-analysis adds the AI triage block: root-cause hypothesis, recommended fix target, failure kind, and the snapshotId tying all artifacts together.

testsprite test result test_3a9f21c7 --include-analysis --output json \
  | jq '{hypothesis: .analysis.rootCauseHypothesis, fix: .analysis.recommendedFixTarget}'

For backend tests the result also carries the captured stdout (apiOutput) and Python traceback (trace): full under --output json (and in failure bundles as result.json / failure.json), a 20-line tail of each in text mode.

Run history

testsprite test result test_3a9f21c7 --history --source cli --since 7d

targetUrl is the address the run started against and stays that way after an environment URL edit; URL userinfo is removed before storage. If a new run still visits an old host, check the saved test steps (Environments).

Flag	Values	Description
--source	cli | portal | mcp | schedule | github_action	What triggered the run
--since	24h, 7d, or ISO timestamp	Earliest createdAt
--page-size <n>	1–100, default 20	Items per page
--cursor <token>	opaque cursor	Continue from a previous page
--env <name>	environment name	Only runs that used this environment
--rerun | --no-rerun	flag	Only reruns, or only fresh runs
--columns <list>	column names	Text output: select or reorder columns
--no-header	flag	Text output: suppress the header row

Triaging a failure

For a failed test, start with the one-screen triage card (status, failure kind, root-cause hypothesis, recommended fix target):

testsprite test failure summary test_3a9f21c7

To pull the full failure bundle to disk:

mkdir -p .testsprite
testsprite test failure get test_3a9f21c7 --out ./.testsprite/failure

--out creates the bundle directory but not its parent; a missing parent exits 5. Without --out, the bundle prints to stdout. The bundle holds the failing step, DOM snapshots as text, test source, and root-cause analysis, all tied to one snapshotId (see Failure Bundle). --failed-only keeps only the failing step and its neighbors (±1) for faster agent context loading.

Pinning to a specific run

test failure get always returns the latest failing run, which moves when a new run arrives. To pin an exact run:

testsprite test artifact get run_5c1d9a2b

Without --out, the bundle goes to ./.testsprite/runs/<run-id>/ (directories created). With --out <dir>, the parent of <dir> must exist. The bundle is immutable, so a concurrent Portal or schedule run cannot overwrite it. Use test failure get <test-id> for interactive triage and test artifact get <run-id> in scripts and CI.

When there is no bundle

The bundle includes the test’s saved code. A frontend test gets saved code when it passes against a public URL; a tunnel (--local) run never saves code. If the test has never passed, or has only run through a tunnel, test failure get and test artifact get exit 4 with reason no_code. A cancelled run exits 4 with cancelled_no_artifacts. In both cases test steps --run-id <run-id> and test failure summary <test-id> still work.

Backend bundles contain saved code and failure analysis (with stdout and traceback when available) but no step log. Use test artifact get <run-id>; on the V3 execution path test failure get <test-id> also returns backend bundles.

Reading test code

testsprite test code get test_3a9f21c7 --out ./checkout_flow.py

Without --out the code prints to stdout; the file’s parent directory must exist. The response includes the language, framework, code, and a codeVersion token that protects testsprite test code put from overwriting a concurrent edit.

Where to Go Next

Running Tests

Trigger runs, wait for verdicts, and understand exit codes

Rerun & Auto-Heal

Re-run a test or let AI repair UI drift

Key Terms

runId, codeVersion, snapshotId, and the rest of the vocabulary

Command Reference

Full flag listing for every CLI command