LibreChat/e2e
Marco Beretta 2bf6ca847a
❓ refactor: Move the Ask-User Question Cards Into @librechat/chat (#18161)
- refactor: Move Ask User Question Cards Into the Chat Package
- fix: Preserve Ask Card Styling Without Suppression Growth
- fix: Provide Live Ask Pauses As Scoped Context Data
- test: Cover the Settled Ask Record Collapse Cascade
- fix: Give the Ask Call Spec Host the Tool Call Chrome
2026-11-14 19:11:57 +01:00
..
benchmarks
benchmarks-fade 🧵 perf: Stop Re-rendering and Re-styling the Thread on Stream Events and Run Boundaries (#18148) 2026-11-14 12:15:48 -05:00
benchmarks-mobile-chat
benchmarks-navigation 🧪 test: Route the Benchmark and BYOM Unroutes Through the Shared Helper (#18085) 2026-11-12 09:47:24 -05:00
benchmarks-reasoning
benchmarks-tree
bombadil
byom 🧪 test: Route the Benchmark and BYOM Unroutes Through the Shared Helper (#18085) 2026-11-12 09:47:24 -05:00
config 🚠 feat: Bridge Registered SDK Agents to Remote Runtimes (#18150) 2026-11-14 10:56:05 -05:00
fixtures 🧾 test: Bill Detached Subagents Their Own Prompt in the Mock Model (#18050) 2026-11-11 10:29:38 -05:00
lighthouse 🧲 fix: Settle Fast Pinned Drops, Defer the Hidden Mobile Header Menu, and Close Untraced E2E Races (#18087) 2026-11-12 09:47:46 -05:00
mcp-apps
perf
recordings
reporters 🎲 ci: Fail Playwright Runs When a Test Passes Only on a Retry (#18051) 2026-11-11 10:31:06 -05:00
rum
screenshots
setup 🏇 feat: Register SDK-Backed Agents for Native Chat (#18114) 2026-11-13 13:02:53 -05:00
specs ❓ refactor: Move the Ask-User Question Cards Into @librechat/chat (#18161) 2026-11-14 19:11:57 +01:00
client-build.test.mjs 🛖 fix: Isolate Sandpack Console Globals From HTML Artifacts (#18093) 2026-11-12 11:42:03 -05:00
config.local.example.ts
jestSetup.js
lighthouse.test.mts 🧲 fix: Settle Fast Pinned Drops, Defer the Hidden Mobile Header Menu, and Close Untraced E2E Races (#18087) 2026-11-12 09:47:46 -05:00
playwright.config.a11y.ts
playwright.config.benchmark.ts
playwright.config.bombadil.ts
playwright.config.deployed.ts
playwright.config.email.ts 🎲 ci: Fail Playwright Runs When a Test Passes Only on a Retry (#18051) 2026-11-11 10:31:06 -05:00
playwright.config.fade-perf.ts 🎞️ perf: Run the In-Flight Label Sweep on the Compositor (#18128) 2026-11-14 10:16:52 -05:00
playwright.config.lighthouse.ts
playwright.config.local.ts
playwright.config.mcp-apps.ts
playwright.config.mermaid.ts
playwright.config.mobile-chat-perf.ts
playwright.config.mock.ts 🧱 ci: Block the PWA Service Worker Across the Mock E2E Profiles (#18063) 2026-11-11 16:22:48 -05:00
playwright.config.navigation-perf.ts
playwright.config.real.ts
playwright.config.reasoning-perf.ts
playwright.config.redis.ts 🎲 ci: Fail Playwright Runs When a Test Passes Only on a Retry (#18051) 2026-11-11 10:31:06 -05:00
playwright.config.rum.ts
playwright.config.tree-parity.ts
playwright.config.tree-perf-prod.ts
playwright.config.tree-perf.ts
playwright.config.ts 🎲 ci: Fail Playwright Runs When a Test Passes Only on a Retry (#18051) 2026-11-11 10:31:06 -05:00
README.md ⏩ ci: Record Playwright Shard Timings Automatically and Run Fifteen Shards (#18115) 2026-11-14 01:25:48 +01:00
timings.mock.json
types.ts

LibreChat e2e

The mock e2e profile is the safest default for generated tests. It starts LibreChat with e2e/config/librechat.e2e.yaml, injects an in-process fake LLM (via LIBRECHAT_TEST_RUN_HOOK), creates an authenticated e2e user, and avoids real provider credentials.

RUM privacy export regressions

Run the real HyperDX/OpenTelemetry SDK export checks without a running LibreChat, database, collector, or model provider:

npx playwright test --config=e2e/playwright.config.rum.ts

The fixture bundles the production RUM adapter, records a synthetic DOM in Chromium, and inspects the actual OTLP requests (including gzip/protobuf replay batches). It checks proxy and direct-token modes, metadata/DOM URL redaction, and fresh and already-recording transitions to a two-factor URL containing a fake login token. Playwright traces are disabled so authentication-shaped test data is not retained in HAR/trace artifacts.

Automatic title regressions

The standard mock profile disables automatic titles. CI runs the title-enabled cases once on the first memory shard. To run them locally after building the app:

E2E_TITLE_CONVO=true npx playwright test --config=e2e/playwright.config.mock.ts \
  conversation-management.spec.ts --grep 'first-turn|final-title polling'

Deployed-instance smoke test

The deployed profile exercises an existing LibreChat deployment without starting another app or database. It uses the deployment's configured model provider and persists a real conversation, so run it only with a dedicated test account in an environment where that traffic is expected.

First, create Playwright storage state by signing in through the deployment's normal login flow:

npx playwright codegen \
  --save-storage=e2e/storageState.json \
  https://librechat.example.com/c/new

Close codegen after sign-in, then run the smoke test:

E2E_BASE_URL=https://librechat.example.com \
  npm run e2e:deployed

The storage-state file contains session credentials. The default path is ignored by Git; do not commit it or include it in test artifacts.

Set E2E_STORAGE_STATE when the auth file is mounted elsewhere. If the account has no default model, set E2E_DEPLOYED_MODEL to the exact configured model label. E2E_DEPLOYED_PROMPT can replace the short default prompt, and E2E_IGNORE_HTTPS_ERRORS=true supports deployments using a self-signed certificate.

The profile deliberately has no global setup, database access, or web server. It verifies the authenticated shell, sends one real prompt, reloads the resulting conversation, and deletes only the conversation created by that run through LibreChat's authenticated API. Keep deterministic provider behavior and destructive database fixtures in the mock profile instead.

Email Delivery

The registered-email flow has a dedicated profile because enabling SMTP changes registration into a verification-required workflow. It starts a disposable local SMTP mailbox, verifies the setup account from the delivered message, and then tests the Settings → Account email change, verification link, notifications to both addresses, and login with the new address:

npm run e2e:email-change

The disabled-mode profile verifies that ALLOW_EMAIL_CHANGE=false hides the Account control, rejects both request and confirmation endpoints, and sends no email:

npm run e2e:email-change:disabled

Override E2E_SMTP_PORT or E2E_MAILBOX_PORT if ports 1025 or 8025 are already in use. The mailbox binds only to 127.0.0.1 and is discarded with the test process.

Stream Stores and Shards

The mock profile uses the in-memory generation stream store by default. To exercise the same browser scenarios through a real Redis job store and pub/sub transport, start Redis on port 6379 and run:

npm run e2e:mock:redis

Memory mode explicitly disables Redis. Redis mode defaults to database 15 with a LibreChatE2E key prefix, and fails closed: the test server pings Redis and verifies that the generation job manager did not silently fall back to memory. Override REDIS_URI or E2E_REDIS_KEY_PREFIX when needed.

Pull request CI runs the complete mock suite in memory mode across fifteen shards, plus a focused Redis transport suite across two. The Redis suite covers streaming fidelity, steering, interrupts, resumptions, HITL approvals, completion, thread folding, model icons, and usage:

npx playwright test --config=e2e/playwright.config.mock.ts --shard=1/15
npm run e2e:mock:redis:transport

The nightly schedule and manual workflow dispatch run the complete mock suite in both stream modes, across fifteen memory shards and ten Redis shards. Every shard keeps one worker so tests do not contend for its authenticated user and database.

CI retries a failed test, twice on the memory shards and three times on the Redis transport suite, so the report carries a trace of each failed attempt and shows whether the failure repeats. A test that passes only on a retry still fails the run (failOnFlakyTests), and the job log and the job summary name it (e2e/reporters/retried.mjs). Local runs do not retry.

CI splits the complete suite by recorded duration rather than test count: each shard lists the suite, and .github/scripts/playwright-shards.mjs plan hands out whole spec files longest-first to the least-loaded shard by recorded seconds. A spec with no recorded time is estimated from its test count, so new specs need nothing.

The timings keep themselves current. On each push to dev, .github/workflows/playwright-timings.yml rebuilds them from the three most recent green runs on dev and saves the result to the Actions cache. The Shard timings job of a later run restores the newest entry once and hands the same copy to every shard, because shards that planned from different timings would run a spec twice or not at all. e2e/timings.mock.json is the fallback whenever no entry can be read, such as the nightly on the default branch or an evicted cache, so it only needs rebuilding if those runs drift apart. Rebuilding it is the command the workflow runs: each shard's report artifact carries a results.json, so download the playwright-report-memory-* artifacts of a few recent runs and rebuild the file from them. A spec measured by several runs is recorded at their mean, which keeps one slow runner from skewing it:

for run in <run-id> <run-id> <run-id>; do
  gh run download "$run" --pattern 'playwright-report-memory-*' --dir "/tmp/e2e-reports/$run"
done
node .github/scripts/playwright-shards.mjs record --out e2e/timings.mock.json /tmp/e2e-reports/*

Route handlers that fetch

A mock spec whose route handler calls route.fetch() takes test and expect from e2e/specs/mock/fixtures.ts instead of @playwright/test. Closing a browser context disposes the responses its handlers fetched, so a handler the app triggered just before the test returned throws Response has been disposed and fails a test that had already passed, which failOnFlakyTests turns into a failed shard. The fixture's context detaches the routes on the context and its pages before Playwright closes it, and what a detached handler throws afterwards is ignored. A context the spec opens itself with browser.newContext() closes through closeContext(context) from the same module.

A spec that removes a route before its test ends calls unroute(page, url, handler) from that module in place of page.unroute(). When page.unroute() removes a page's last route, Playwright stops intercepting there and continues the requests its handlers were still working on, so a handler in the middle of one throws Route is already handled! and fails the test. unroute first adds a route that matches nothing, which keeps the page intercepting: handlers already running answer their requests, and later requests reach the server.

Harness servers and idle connections

Playwright sends route.fetch(), the request fixture and page.request through one keep-alive agent per worker that never closes an idle socket and ignores a server's Keep-Alive: timeout=N hint. A Node server closes an idle socket after keepAliveTimeout plus keepAliveTimeoutBuffer (6 s by default), so a request that reuses the socket at that moment fails with read ECONNRESET or socket hang up. The harness therefore runs LibreChat with HTTP_KEEP_ALIVE_TIMEOUT_MS=0, and every fixture server the mock profiles start as a webServer passes its server through keepIdleConnections() from e2e/setup/keepalive.js. A new fixture server that specs reach from Node does the same. fixtures.spec.ts checks that each harness server answers without an idle deadline.

Property-based browser testing

Bombadil explores randomized sequences across the core chat loop, message branches, parallel multi-conversation responses, model changes, reloads, and sidebar conversation lifecycle operations:

npm run e2e:bombadil

Set BOMBADIL_TIME_LIMIT for longer local or scheduled runs. Failures leave a reproducible trace under e2e/.generated/bombadil-output; rerun it with:

BOMBADIL_REPRODUCE=e2e/.generated/bombadil-output npm run e2e:bombadil:run

Reproducing a real violation is expected to fail the Playwright test. Before a new run overwrites the active output, the harness archives it under e2e/.generated/bombadil-history/. Reproduction can diverge when streaming timing changes; Bombadil reports that explicitly.

The harness uses the credential-free mock-LLM profile, so exploration never sends billable provider requests.

CI runs the broad property exploration for five minutes in the non-blocking Bombadil Property Exploration workflow. If a property fails, download the bombadil-reproduction-* artifact into e2e/.generated/bombadil-output/, then reproduce it locally:

BOMBADIL_REPRODUCE=e2e/.generated/bombadil-output npm run e2e:bombadil:run

The accompanying bombadil-diagnostics-* artifact contains the captured CI log, Playwright HTML report, and Playwright test results. A Bombadil failure produces a workflow warning but does not block merge.

The default instruments inline JavaScript only because instrumenting LibreChat's full Vite bundle can exceed Bombadil's driver timeout during stateful runs. Set BOMBADIL_INSTRUMENT_JAVASCRIPT=files,inline for shorter coverage-guided experiments.

The branch reload, fork submission, model/conversation, HITL pause/resume, and mid-run steering lifecycle properties can be run independently:

npm run e2e:bombadil:branch-reload
npm run e2e:bombadil:fork-lifecycle
npm run e2e:bombadil:model-lifecycle
npm run e2e:bombadil:hitl
npm run e2e:bombadil:steering

These focused commands are diagnostic properties: they exit nonzero when they reproduce a product invariant violation. Reproduce a focused trace with its matching :run script and output directory, for example:

BOMBADIL_REPRODUCE=e2e/.generated/bombadil-output-hitl npm run e2e:bombadil:hitl:run

HITL drives a real ask_user_question checkpoint through the answer/resume controller, reloads while the question is paused, answers it once, and reloads the completed conversation. Steering submits an in-flight steer during a slow MCP-backed run, checks that it moves exactly once from the composer anchor into the response at the tool boundary, and reloads the applied state. The model lifecycle property is the passing control. The branch reload and fork properties preserve their minimal failing traces.

Recording Tests

Use Playwright codegen when you want to turn an exploratory browser session into a draft test:

npm run e2e:record

That command builds the app, starts the LibreChat test server (with an in-process fake LLM) when needed, writes e2e/storageState.json, and opens Playwright codegen at /c/new. The npm script uses http://localhost:3333 so it does not collide with a normal dev server on 3080. Raw recordings are written to e2e/recordings/ and ignored by git.

For a real local LibreChat config instead of the mock profile:

npm run e2e:record:local

Useful direct options:

node e2e/setup/record.js --url=http://localhost:3080/c/new
node e2e/setup/record.js --profile=local --no-output
node e2e/setup/record.js --auth-only
node e2e/setup/record.js --output=e2e/recordings/settings-draft.spec.ts

LLM-Assisted Loop

  1. Start npm run e2e:record.
  2. Let the LLM use Computer Use to operate the headed Playwright browser.
  3. Stop codegen after the workflow is captured.
  4. Move the useful parts from e2e/recordings/ into a committed spec under e2e/specs/mock/.
  5. Replace brittle generated selectors with role, label, text, or data-testid locators.
  6. Add assertions that prove the behavior, not just the clicked path.
  7. Run the finished spec with npm run e2e:mock -- <spec name>.

Generated recordings are a draft, not the final test. The committed version should use the shared helpers in e2e/specs/mock/helpers.ts where possible, wait on network or visible UI state instead of fixed sleeps, and keep test data deterministic.

PR screenshot pilot

e2e/screenshots/playwright.config.ts is an opt-in evidence capture lane, not a pixel-baseline suite. It runs the real app with the mock profile and a private ephemeral MongoDB, then captures welcome screens at desktop/mobile sizes in both themes. Desktop captures also include temporary chat and the settings dialog. No model request is made. Motion is reduced, so these stills do not prove transitions or streaming behavior.

Build and run each revision in its own clean worktree with its own locked install:

npm ci
npm run frontend
npx playwright install chromium
E2E_CAPTURE_SHA=$(git rev-parse HEAD) \
E2E_CAPTURE_DIR="$PWD/e2e/.generated/evidence-before" \
E2E_USE_MEMORY_MONGO=true \
E2E_BASE_URL=http://127.0.0.1:3333 \
npx playwright test --config=e2e/screenshots/playwright.config.ts

For a historical revision without this lane, copy the two e2e/screenshots/*.ts files into its worktree, leaving its application code and lockfile unchanged. Run the same scenario source on both revisions. Use a new output directory for every attempt: existing images are never overwritten. Run revisions sequentially, or give every fixture service its own port using the mock profile's E2E_*_PORT settings. Do not reuse a running development server or a real user's database.

Each PNG has a JSON sidecar containing the revision, browser version, viewport, theme, scenario hash, lockfile hash, built HTML hash, and image hash. Visible images must decode, fonts must finish loading, greeting springs must settle, and consecutive captures must match. A passing run and human inspection are both required before treating the pair as reviewed evidence; sidecars from a failed run are not complete evidence. Compare matching filenames between revisions. Pin the browser and font environment as well as the app revisions.

Keep storage state, traces, logs, and session data private. Check that the PNGs contain only intended synthetic test data before uploading with the attachment-capable gh described in the PR template, then read back the PR body to verify real asset URLs. Mark any pending visual review explicitly. Do not commit the images. A new surface has no before state; label it accordingly rather than substituting another screen.

Running-chat rename rollout

interface.runningChatRename defaults to true. Set it to false during rolling upgrades from a version without title ownership. After every API replica supports title ownership, drain all title-generation tasks started on older replicas, then remove the override. The client also requires the server's title-ownership capability; an old replica cannot advertise it.

With the fence off, running rows and unowned placeholder titles cannot be renamed. Other menu actions remain available. Settled non-placeholder chats retain Rename. The isolated mock profile enables the option because it contains no older replicas.