* test(mobile): fake the clock and shrink PNGs in engine tests, trim duplicates
engine.test.ts 6.4s -> 0.4s: retry backoff and transition budgets on fake timers, small screenshots. Merge or drop tests restating constants or covered elsewhere; prose asserts in errors.test check the headline and recovery only.
* test(mobile): delete the simulator suite no workflow runs
E2E_AGENT_DEVICE_SIMULATOR was never set in CI; the mobile benchmark drives real simulators and emulators. Drops test:simulator and the integration project.
* test(mobile): cover duplicate refind, clipboard step label, unreadable secure screenshot
* test(eas): restore Date.now in afterEach, unique run ids, drop constant asserts
* test(github): drop the factory tautology and markdown asserts core owns
* test(mobile): keep the videoTouches and interactive snapshot option checks
* test(mobile): name the openApp permission refusal by what it checks
* fix(e2e): load TypeScript with an oxc loader instead of tsx
* fix(e2e): close loader edge cases: sourceless CommonJS, export maps without extensions, unreadable tsconfig, chained source maps, Windows require paths
* fix(e2e): refuse Bun and Deno runtimes, name oxc load failures, follow Yarn PnP export targets
* perf(e2e): skip the syntax check for files whose only @ is in a package import
* fix(e2e): allow type-only exports in .cts, warn from runner only, Bun check first, docs
* fix(e2e): workers warn about tsconfigs runner did not, FILE: urls, foreign infra errors, review tests
* fix(run): keep --last-failed reruns honest: failure-limit skips, carried debts, kept evidence
--last-failed now reruns tests --max-failures skipped. A test or failed
suite hook another filter leaves out stays owed in the report's new
run.carried until a rerun runs it; the github fold shows carried rows and
errors and stays red while anything is carried. A rerun prunes artifacts/
to what the previous report names and writes its attempts under
artifacts/rerun-<n>/ instead of wiping the tree and reusing attempt paths.
* fix(run): carry a setup's hook debt when no selected test needs it
The report has no row for an unneeded setup, so carryForward now asks the
collection whether a test still exists instead of the report's rows.
* fix(run): reject malformed artifacts in --last-failed reports, tighten rerun docs
Distinct ids in the carried fixture row.
* fix(run): keep carrying tests of a file a narrowed rerun could not collect
* fix(github): a carried interrupted test reads interrupted, not failed, in the folded page
* fix(report)!: count interrupted tests apart from failures
* fix(report): tell a failed test from its last failed attempt, whatever the cut retry left
* fix(github): long marker fields keep a digest so distinct keys never share a comment
* docs(changeset): note one-time duplicate for long marker fields
Drops NPM_TOKEN from release.yml; each package grants release.yml publish
rights via an npm trusted publisher. Release job moves to ubuntu-latest
because npm rejects OIDC from self-hosted runners (Blacksmith included).
publishConfig.provenance pins removed so provenance attaches automatically
once the repo is public. Setup recipe documented in CONTRIBUTING.md.
* chore(npm): e2e ships the repo README, revised keywords for every package
* test(skill): follow the skill copy into prepare-build.ts
* fix(npm): keep README links relative on GitHub, rewrite them for the npm copy
* feat!: the app is declared on its target; engines only drive it
The app under test moves from the engine's options to `targets: [{ engine,
app }]`: `url`, `bundleId`, `appPath`, `identity`, `environment`,
`launchArguments`, `permissions`, `command`, and `readyUrl`. `web()` and
`mobile()` refuse every moved option in one `INVALID_CONFIG` naming the
target block; `defineEngine({ app })` is gone and engines check what their
platform needs in `validateApp`. `requires: ['browser' | 'native-app']` is
derived from the target's app. Replay cache and session keys are byte for
byte what they were, pinned by target-identity.test.ts.
Dependency services are not part of this version: `services` anywhere, and
`web({ services })`, fail saying so. `app.command` on port 0 without a
command, and `reuseExisting` on a free port, are refused. An `e2e mcp`
open cancelled or disconnected mid-start stops what it started, and a hung
shared start is abandoned once no session waits on it.
* fix: review follow-ups: cancelled opens, abandoned starts, null targets, stale docs
* refactor: drop the moved-option messages; old app keys are unknown keys
* test: give the replaced-nodes suite a budget that survives a loaded runner
* test: size the replaced-nodes hook budget to its seven tests
* chore(release): versioned releases publish to latest, ship the github reporter with the stable runner
Versioned releases went to the beta dist-tag while latest stayed on a July placeholder, so a bare npx e2e init installed 0.1.0. The release script now passes no --tag and publishConfig.tag reads latest on every package.
@e2e-dev/github had no pending changeset; its only published builds peer on one exact canary runner and refuse a stable e2e. A patch changeset ships it alongside 0.15.0 with the wide peer range.
Quickstart and the skill setup reference install plain e2e instead of @canary. CONTRIBUTING and AGENTS describe the latest channel.
* chore(changeset): word the github reporter entry for the changelog
* fix(agent): recover semantic timeouts with masked pixels
Distinguish unavailable semantics from an empty tree so a proven fresh screenshot can remain useful without supporting an absence judgment. Reserve time for independent masking, reject stale references, and stop trace replay or recording after evidence degrades.
Respect judgment vision settings and attempt pixel taint. Keep device capture closed when accessibility bounds cannot prove masking. Report pre-model capture failures without inventing a revision or verdict.
* fix(agent): isolate fallback capture and evidence ownership
Reserve document node IDs before readers run and bound the entire semantic transaction, including metadata and handle readback. Quarantine late handles and publish each capture only while its page and generation remain current.
Carry semantic or pixel evidence through the internal session and agent paths. Preserve completed cache captures for one executor handoff, retire semantic presentation history on fallback, and keep public interfaces extendable.
* test(agent): align capture assertions with cache handoff
Count the reused cache observation once and wait for the fixture delayed control before asserting an unchanged screen after batched actions.
* docs(agent): clarify fallback without act vision modes
* test(agent): cover requested pixels after semantic fallback
* test(playwright): wait for recorded paint in lifecycle coverage
* test(playwright): flush a frame before checking video paint