## Summary
Enable additional lint rules that catch unsafe optional-chain
assertions, inherited-property iteration, anonymous symbols, and unsafe
external links.
The existing violations now use explicit values and own-property checks,
so the rules can prevent those patterns from returning.
## Summary
The server half of hosted webhooks: the public ingress endpoint,
signature verification, the delivery pipeline (Postgres partitioned
storage + ClickHouse for ordering), the in-app partition manager, the
HTTP API, and the dashboard (Deliveries, Endpoints, and the in-app test
console).
The public SDK and docs half is #4537. That PR carries the user-facing
API (`webhook()`, `chat.event` / `chat.channels`, the
`@trigger.dev/slack` connector) and builds on the shared
`@trigger.dev/core` schemas that ship here.
## Shipping behind a flag
A `WEBHOOK_ENABLED` env var (default off) gates the public ingress route
and the engine worker plus partition cron, so merging and deploying this
changes nothing in production until it is flipped on per environment.
The dashboard is separately gated per org by the `hasWebhooksAccess`
feature flag.
## Note on packages
This PR includes the `@trigger.dev/core` schema additions the server
compiles against, but carries no changeset. Core is not consumed
independently of the SDK, so it is released together with the SDK via
#4537. Keeping its changeset off `main` means no release cut from `main`
publishes it early.
## What & why
This is the system behind the Dashboard Agent — an assistant that
answers questions about a project's runs, errors, queues, deploys and
health, and can investigate failures end to end.
The agent runs as a chat.agent task in its own Trigger project. It has
no access to the main database or ClickHouse; all platform data is read
through the public API using a delegated, read-only user token.
Everything here is behind `canAccessDashboardAgent` and inert with the
flag off. The UI that mounts the panel lands in #4529.
## Stack
`#4418` (this, base) ← `#4529` UI ← `#4525` Watch ← `#4516` storybook
gallery. The scenario/contract reference for the whole stack is
`internal-packages/dashboard-agent/GUIDEBOOK.md` (it lands on the Watch
branch): it states, per feature, what makes each thing happen and where
that is decided.
## What's inside
**Agent runtime and tools** — `internal-packages/dashboard-agent`:
prompt, tool set (API reads, TRQL query, docs, navigation,
evidence/investigations, repo source), conversation compaction, a
prompt-prefix token budget pinned by snapshot test, and sampled
LLM-judged turn evals. The package cannot import webapp server code,
which is what makes the "no DB access" claim structural rather than a
convention.
**Contracts** — `internal-packages/dashboard-agent-contracts`:
`trigger://` URIs, intents, and the block envelope every rendered card
travels in.
**Conversation store** — `internal-packages/dashboard-agent-db`: drizzle
over postgres-js in its own `trigger_dashboard_agent` Postgres schema,
plus one additive migration.
**Auth boundary** — the user-actor token gains an optional environment
claim; one guard (`userActorEnvironment.server.ts`) enforces it so
routes don't each re-derive the rule. Token minting, cap ceiling, and
the RBAC fallback path for self-hosted.
**Transport** — webapp resource routes that mint the token and proxy
each turn, and SDK-side mid-turn reconnect.
**Public API the agent reads through** — orgs, projects, environments,
runs, queue metrics, workers, a run's commit metadata, repo snapshot,
reports, and `POST /api/v1/query`.
**Reports** — the health report's layout is declared once and shared by
the card, the markdown surface and the JSON/MCP surface, so the same
report reads the same in the dashboard, the terminal and an editor.
**Block renderers** — the report and investigation cards the flows above
already emit (`app/components/dashboard-agent/`). The panel that hosts
them, and the rest of the chat UI, is #4529.
**Query safety and CSP** — see below.
## Key decisions
- **The agent is a separate Trigger project, not webapp code.** It reads
platform data over the public API with a delegated user-actor token
whose `cap` ceilings it to read scopes. No Prisma, no ClickHouse, no
webapp imports.
- **The PAT-only auth helper now refuses user-actor tokens.** This is an
intentional behavioral change: its callers consume only a bare userId
and do not enforce delegated-token capabilities. Actor-aware routes
continue through the scoped route builders instead.
- **RBAC fallback builds a delegated token's ability from its own cap**,
never the blanket ability a PAT gets (read-only when the token declares
none). Without this, the agent's read-only cap would buy a write JWT on
self-hosted.
- **Org creation checks RBAC only for user-actor tokens, and only after
the env gate**, so an install with `ORG_CREATION_API_ENABLED` off
returns 404 rather than 403, and an ordinary PAT never consults an
ability the route has no org to scope. Both orderings are pinned by
test.
- **The query path is read-only in depth.** TRQL rejects write
statements at the grammar level (they don't parse, rather than being
filtered), ClickHouse runs with `readonly=1`, and the org/project/env
filters are injected server-side from the credential — the request body
cannot widen scope. An unparseable query denies instead of falling
through to the permissive resource.
- **Document-wide img-src CSP.** Remote images are an
outbound-request/exfiltration surface, so the policy permits only
own-origin/data/blob, the required SSO avatar hosts, and the favicon
endpoint. Operators can add exact origins through CSP_IMG_SRC_ALLOWLIST;
wildcard hosts and bare schemes are intentionally not allowed.
- **The chat transport reconnects on a mid-turn EOF**
(`@trigger.dev/sdk`). A body that ends without a turn-complete is
terminal only when the server says `X-Session-Settled: true`; otherwise
the transport resubscribes from `lastEventId` with bounded backoff, and
any record re-earns the budget. Previously a closed long-poll window or
a proxy restart left the reply stuck as if still generating.
- **Conversations live in their own datastore**, schema-scoped and
foreign-key-free (it references `organizationId`/`userId` by id, because
in cloud it is a different database). It is a display read-model for the
History tab and transport resume; `chat.agent`'s object-store snapshot
remains the model's source of truth.
- **Deterministic first.** Reports and health checks contain no LLM —
they are computed from the same data the dashboard shows, and the model
only narrates and links them. That is what makes a number in an answer
auditable.
## Testing
- 63 new test files, run with `pnpm run test --filter webapp` and
per-package vitest. Heaviest coverage on the auth boundary
(`userActorPatOnlyBoundary`, `userActorTokenClaimsAndScopes`,
`contextlessPatRoutes`, `rbacFallbackBranch`), TRQL read-only, the
report layout, and the SDK reconnect.
- The agent package has a separate eval lane (`pnpm run test:evals`,
`vitest.eval.config.ts`) that hits the real model, so it never runs in
`pnpm test`.
- Live-tested against a local stack scenario by scenario; the GUIDEBOOK
lists the condition each behaviour is expected under, which is what
those runs were checked against.
## Changelog
`.server-changes/dashboard-agent.md`, plus changesets for
`@trigger.dev/core` (report schemas), `@trigger.dev/sdk` (chat
reconnect) and the CLI's `mint-token` help text.
Most deploy-flow API calls shared the general per-environment rate limit
bucket with all of that environment's runtime traffic, so an org with
heavy API usage could intermittently 429 its own deploys; the
`/api/v*/deployments` endpoints themselves were fully exempt from rate
limits as a stopgap
([#2774](https://github.com/triggerdotdev/trigger.dev/pull/2774)), which
promised a dedicated limiter as the follow-up. This is that follow-up:
the whole deploy-flow group now runs on its own budget, separate from
runtime API limits.
### Design
A new `deploymentRateLimiter` covers every endpoint the deploy flow
depends on: the `/api/v*/deployments` group, the env API key exchange
(`/api/v1/projects/:ref/:env`), build-time env var resolution and sync
(`/envvars`, `/envvars/:slug/import`), preview branches,
`/api/v1/remote-build-provider-status` and `/api/v1/artifacts`. The
general API limiter whitelists the same shared path list, so exactly one
limiter applies to each path and the two can't drift apart.
Buckets are keyed per environment for environment API keys and per token
for the PAT-authenticated phase of a CLI deploy (whoami, key exchange,
branches). The deploy budget is controlled via the
`DEPLOYMENT_RATE_LIMIT_*` env vars.
## Problem
The webapp's HTML references content-hashed `/build` assets, and each
running
instance contains exactly one build and returns 404 for asset hashes it
doesn't
have. During a rolling deploy a client can hold HTML from one build
while a
request for one of its assets is served by an instance on a different
build →
missing styles or a failed chunk load.
## What this does
On a `/build` stylesheet/script/chunk load failure, the client does a
**bounded
full-document reload** (at most 2 per 5 minutes, tracked in
`sessionStorage`) so
the page reloads onto a single consistent build. That's the whole
mechanism — no
polling, no `fetch` interception, no blocking overlay, no form
snapshotting.
- `apps/webapp/app/components/StaleAssetRecovery.tsx` — authored as a
typed,
lint-checked function and serialized to an inline script via
`.toString()` (so
the logic is real, reviewable code, not an opaque string), injected
before
`<Links />`, production only.
- Detection: capture-phase `error` listener for
`<link>`/`<script>`/modulepreload
failures under `/build/`, plus an `unhandledrejection` guard for
dynamic-import
failures.
- Guards: once-per-page re-entrancy guard, the bounded reload budget,
and a
`navigator.onLine` check so it never reloads into an offline error page.
- Unit tests in `StaleAssetRecovery.test.ts`.
## Relationship to #4260
Replaces the recovery introduced in #4260 (reverted in #4280) with a
much
smaller, reload-only approach — the previous version intercepted `fetch`
and
could turn a data request into a navigation, and showed a full-screen
overlay on
any asset error; this drops both.
## `/build-version` compatibility shim
`apps/webapp/server.ts` adds a tiny `GET /build-version` endpoint (build
id only,
`no-store`). A previously-deployed client build polls it after an asset
failure
and reloads once it sees a newer build, so those older tabs recover in
one reload
instead of getting stuck. Temporary — safe to remove once older clients
have
cycled out. It deliberately does **not** re-add an `X-Build-Id` response
header.
## Also
Restores the `.server-changes` writing guidance in
`.claude/rules/server-apps.md`
(reverted alongside #4260).
## Self-hosting note
Recovery is most reliable when your load balancer keeps a client on one
instance
for the duration of a deploy (short session stickiness) — the reload
then lands
on a consistent build in one hop.
Standard `git revert` of #4260.
Its client-side stale-asset recovery is net-negative during normal
deploys:
- The `fetch` interception treats any `?_data=` request (Remix loader
**and** action traffic) as a navigation and, on a build-id mismatch,
`location.assign`es the tab to the fetched URL — an open dashboard tab
can be hard-navigated to a raw data URL during a rolling deploy, losing
unsaved input.
- Any transient `/build` asset error (a network blip, an extension, an
unrelated failed dynamic import) blanks the page behind a full-screen
overlay for ~60s before offering a manual reload.
- It serialized form field values to `sessionStorage` to restore them
across the reload.
This returns the webapp to the pre-#4260 baseline as a fast, low-risk
step.
Follow-ups (separate PRs):
- a minimal reload-only recovery to replace this,
- restore the unrelated `.claude/rules/server-apps.md` docs tidy-up from
#4260 (via cherry-pick),
- a load-balancer stickiness change addressing the root cause.
### Problem
Webapp HTML references content-hashed /build assets, and each Docker
image contains exactly one build with a hard 404 for unknown hashes.
During a rolling deploy, a client holding HTML from the old build may
request old asset hashes from a replica running the new image, causing
missing styles or failed chunk loads.
The page should recover automatically once a compatible build becomes
available, without reload loops or unnecessary interruptions during
normal deployments.
### What changed
- Build changes alone do nothing — no polling, no automatic reloads.
- If a CSS or JavaScript asset fails to load, a recovery overlay is
shown immediately.
- While the server still reports the same build, the client polls for a
newer build using exponential backoff (up to ~60s). As soon as a newer
build is detected, the page reloads automatically.
- If no newer build appears within the timeout, recovery falls back to a
manual Reload action.
- If recovery still fails after the automatic reload, the client stops
retrying and displays a final recovery screen instead of entering a
reload loop.
- Recovery preserves form values and scroll position across the
automatic reload.
## Summary
posthog-js sent product analytics to PostHog Cloud directly from the
browser. This points `api_host` at a same-origin `/ph` path that
forwards to PostHog Cloud EU server-side, following PostHog's standard
first-party reverse-proxy setup.
## How it works
A resource route forwards each request server-side, splitting by path:
`/ph/static/*` and `/ph/array/*` go to the asset host, everything else
(analytics events, feature flags) goes to the ingest host. It rewrites
the `Host` header, strips the `/ph` prefix, and streams the response
back. Only PostHog's own cookies are forwarded, so the app session
cookie stays first-party. Upstream hosts default to PostHog Cloud EU,
overridable via `POSTHOG_INGEST_HOST` / `POSTHOG_ASSETS_HOST`.
It also sets `cross_subdomain_cookie` so a single PostHog session is
shared across the marketing site and app.
Verified locally: static assets return 200 from the EU asset host, and
analytics events return 200 through the ingest host.
Once this is merged, oxlint is at a pretty sensible baseline.
**Enable `no-unused-vars`, `typescript/consistent-type-imports`, and
`import/no-duplicates` lint rules**
Turns on three previously-disabled oxlint rules across the monorepo and
fixes all violations:
- **`no-unused-vars`** – enabled as an error with standard ignore
patterns: unused function arguments are ignored by default (`args:
"none"`), variables/caught errors/destructured array elements prefixed
with `_` are allowed, and rest siblings are permitted.
- **`typescript/consistent-type-imports`** – enforced as an error; all
type-only imports now use the `import type` syntax.
- **`import/no-duplicates`** – enforced as an error; duplicate import
statements from the same module have been merged.
The remaining commits clean up the violations found across the codebase:
removing unused variables/imports/type aliases, adding `_` prefixes to
intentionally unused bindings, fixing duplicate imports, and converting
value imports to `import type` where appropriate.
## Summary
Adds Billing Limits to the webapp.
Customers can set a monthly spend cap. When usage crosses the limit,
billable environments enter a grace period. If the limit is not resolved
before grace expires, new triggers are rejected until the organization
increases or removes the limit.
## Summary
- `internal-packages/rbac/src/index.ts` — in `LazyController.load()`'s
catch block, throw an Error when `process.env.REQUIRE_PLUGINS === "1"`
instead of silently falling back. The throw is captured into the lazy
controller's init promise, so it surfaces on the first method call.
- `apps/webapp/app/routes/healthcheck.tsx` — `await
rbac.isUsingPlugin()` after the DB ping. With `REQUIRE_PLUGINS=1` and a
failed plugin load, the throw surfaces here and the healthcheck returns
500 → readiness probe fails → rollout is rolled back. Noop for
self-hosters.
- `.server-changes/require-plugins-fail-fast.md` — server-changes entry.
- `internal-packages/rbac/src/require-plugins.test.ts` — 4 unit tests
covering loader branching: unset → fallback, `=1` → throw,
`forceFallback: true` wins, only exactly `"1"` enforces.
- `internal-packages/testcontainers/src/webapp.ts` — adds
`requirePlugins?: boolean` to `StartWebappOptions`. Implies
`forceRbacFallback: false`.
- `apps/webapp/test/healthcheck-require-plugins.e2e.test.ts` — e2e
closes the loop: spawns a real webapp, hits `/healthcheck` via HTTP,
asserts 500 with `REQUIRE_PLUGINS=1` and 200 without.
## Motivation
Today the RBAC plugin loader catches any plugin-load failure (missing
module, broken transitive dep, init throw) and silently returns the
default fallback implementation. This is the correct behaviour for
self-hosters who don't ship the plugin — but it's dangerous in
deployments where the plugin is expected to load: an
accidentally-missing or broken plugin would silently disable
enforcement.
`REQUIRE_PLUGINS=1` makes the loader fail loudly in those deployments.
The variable name is intentionally plural and generic — future plugin
contracts (audit logs, SSO) can read the same flag without renaming.
Closes
[TRI-9852](https://linear.app/triggerdotdev/issue/TRI-9852/require-plugins1-fail-fast-for-required-plugin-loads).
## Test plan
- [x] `pnpm run test --filter @trigger.dev/rbac` — 38/38 tests pass,
including the 4 new loader tests
- [x] `pnpm run typecheck --filter webapp` — passes
- [x] `pnpm run typecheck --filter @trigger.dev/rbac --filter
@internal/testcontainers` — passes
- [x] e2e test added (`healthcheck-require-plugins.e2e.test.ts`) — CI
runs it via `e2e-webapp.yml`. Couldn't run locally (no Docker daemon
up); CI has Docker provisioned.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
## Summary
On a busy webapp the trace/log/metric ingestion path emits several
`info` logs per insert batch, which makes up the bulk of the service's
log output. This moves that per-batch chatter to `debug` and adds an
opt-in to drop successful HTTP access logs, cutting log volume with no
loss of error signal.
## Details
The per-batch ClickHouse insert logs, the flush scheduler's concurrency
adjustments, and the event-loop utilization sample (already exported as
a metric, so the log line was redundant) now log at `debug`. Error and
warning logs are untouched.
New `HTTP_ACCESS_LOG_DISABLED=1` env var: when set, the HTTP access
logger skips successful (2xx) requests while still logging non-2xx
responses. Defaults off, so existing deployments are unchanged.
Make the Express server's `keepAliveTimeout` configurable via
`HTTP_KEEPALIVE_TIMEOUT_MS`. Default preserved at 65000 ms — no behavior
change if unset.
## Summary
Stamp every Sentry event with the signed-in user and the tenant (org /
project / env) the request belongs to, so "Users Impacted" counts
distinct humans and events become filterable per tenant.
**Design after review (current):**
- `user.id = real user cuid` (from `requireUser`). "Users Impacted"
counts humans, not tenants.
- Tenant context (org / project / env slugs, IDs, env type) moves
entirely onto tags: `org_slug`, `project_slug`, `env_slug`, `org_id`,
`project_id`, `project_ref`, `environment_id`, `env_type`, plus
`impersonating` when set.
- Backed by an `AsyncLocalStorage` scope established at the HTTP entry.
Each entry point fills what it knows; loaders enrich the same scope with
what they already have.
**Zero new database queries.** The middleware does a regex match only.
Dashboard loaders that already query Prisma gain a couple of extra
selected columns; nothing new round-trips.
## How it's wired
- **Express middleware (`tenantContextResolver.server.ts`)** — parses
the URL with a regex and always opens an ALS scope. Populates whatever
subset of slugs is present: `/orgs/:o` → just `orgSlug`;
`/orgs/:o/projects/:p` adds `projectSlug`; the full triple adds
`envSlug`. Non-tenant paths get an empty scope so loaders can still
enrich.
- **`_app/route.tsx`** — already calls `requireUser`. Adds
`tenantContext.enrich({ userId: user.id })` for every authenticated
dashboard request. No new query.
- **Env layout loader (`_app.orgs.$o.projects.$p.env.$e/route.tsx`)** —
its existing `prisma.project.findFirst` gains two columns in `select`
(`externalRef`, `organization.id`). After it picks an env, calls
`tenantContext.enrich({ orgId, projectId, projectRef, envId, envType
})`. Same query, +2 columns.
- **API path (`apiBuilder.server.ts`)** — wraps every handler in
`tenantContext.run(tenantContextFromAuthEnvironment(authenticationResult.environment),
…)`. The mapper pulls `userId` from `env.orgMember?.userId` (already
selected by `authIncludeBase` — no schema change). Covers
`createLoaderApiRoute`, `createActionApiRoute`, and
`createMultiMethodApiRoute`.
- **Event processor (`sentryTenantContext.server.ts`)** — registered in
`entry.server.tsx` so it lives in the Remix bundle and shares the same
`tenantContext` ALS instance as the middleware and loaders. Stamps
whatever's present; nothing forced.
## Example events from local verification
| URL | `user.id` | Tags |
|-----|-----------|------|
| `/orgs/:o/projects/:p/env/:e/...` | real user cuid | `org_slug`,
`project_slug`, `env_slug`, `org_id`, `project_id`, `project_ref`,
`environment_id`, `env_type` |
| `/orgs/:o/settings` (non-env-scoped) | real user cuid | `org_slug`
only |
| API request with `orgMember` | `orgMember.userId` | full tenant set |
| API request without `orgMember` | (unset) | full tenant set |
## Trade-offs
1. On env-scoped pages, errors that fire before the env layout loader's
enrich callback runs get slugs + `user.id` but not the tenant IDs /
`env_type`. Realistic errors deep in async work get the full set. (Same
race as before, narrower window now that slugs/`user.id` are populated
up-front by the middleware and `_app` enrich.)
2. API requests where the environment has no `orgMember` get tenant tags
but no `user.id`. Those events still show in the issue but don't
contribute to "Users Impacted".
## Out of scope (deferred)
Background workers (`redis-worker`, `schedule-engine`) and socket
handlers. Those entry points don't set `tenantContext.run` yet — their
events ship without tenant attribution until each is wired in a
follow-up.
## Tests
31 unit tests across 4 files. New tests notably cover:
- `parseTenantPath`: org-only, org+project, and full-triple URL
variants.
- `tenantContext.enrich`: in-place patch, no-op outside `run()`,
concurrent-scope isolation, empty-scope + enrich pattern (for non-tenant
pages).
- `tenantContextFromAuthEnvironment`: with and without `orgMember` —
verifies the API path's `user.id` mapping.
- `addTenantContextToEvent`: empty scope, userId-only, slugs-only, full
enrichment, conditional tag emission, preservation of prior `event.user`
fields.
## Test plan
- [ ] `pnpm run typecheck --filter webapp`
- [ ] `pnpm run test --filter webapp -- test/tenantContext.test.ts
test/sentryTenantContext.test.ts test/tenantContextResolver.test.ts
test/tenantContextFromAuthEnvironment.test.ts`
- [ ] Local manual: with `SENTRY_DSN` set, hit a dashboard URL and an
API route, confirm the captured events carry `user.id` + the expected
tag set in Sentry.
- [ ] After ship: confirm "Users Impacted" on a real Sentry issue
reflects distinct users (not tenants).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Pool Redis connections for non-blocking ops (ingestData, appendPart,
getLastChunkIndex)
using a shared singleton instead of new Redis() per request. Use
redis.disconnect()
for immediate teardown in streamResponse cleanup. Add 15s inactivity
timeout fallback.
Fix broken request.signal in Remix/Express by wiring Express
res.on('close') to an
AbortController via httpAsyncStorage. All SSE/streaming routes now use
getRequestAbortSignal() which fires reliably on client disconnect,
bypassing the
Node.js undici GC bug (nodejs/node#55428) that severs the signal chain.
* feat(webapp): add support for running web services (api, engine, webapp) in cluster mode for better perf
* cleaned up signal handling and resolved some valid 🐇 issues
* Sentry WIP
* Configure sentry for uploading and releasing during the publish webapp step
* Delete source maps after uploading
* Forward logger.error calls to sentry through Logger.onError
* Couple tweaks to the dockerfile
* remove registry proxy
* remove --self-hosted flag
* automatically set network build flag
* update syncEnvVars debug log
* improve switch command
* always display deploy errors if they exist
* fix stuck deploy command after finalize error
* webapp-driven deploys, multi-platform support, lots of fixes
* add worker deployment migration
* rename image platform env var
* only try to sync parent env vars for preview deployments
* add KEEP_TMP_DIRS
* supervisor: docker api version lock, auth, multi-platform
* set image ref on create, validate digest
* use metadata for digest, fix local multi-platform builds
* print git meta branch before commit
* improve push and load flag handling
* make runs after local builds compatible with load and push
* small improvement for platform overrides
* add image platform to dequeued message
* remove deprecated init request body fields
* fix fail deployment id param
* remove build debug logs
* pass report merge with no tests
* structured run debug logs
* add required env var for tests
* should not be an error log
* add changeset
* Upgrade posthog-node to clear axios vulns
* Upgrade @slack/web-api to use a secure version of axios
* Update parse-duration to fix security vulns
* Mitigate against the ws DoS vuln
* Upgrade body-parser to 1.20.3 in the webapp
* Upgrade react-use to 17.5.1 to remove the fast-loops transitive dep
* Remove unused babel dev deps and config file that's no longer used
* Upgrade prismjs to 1.30.0 and bundle parse-duration now
* upgrade express to 4.20.0 to fix issue with XSS when redirecting
* Upgrade @conform/zod to 0.9.2
* bump worker version
* Suggested glossary for the RunEngine, TBC
* Removed BatchTaskRun changes from this branch, they were done in main
* Set the BatchTaskRun status to completed when all runs are completed
* When dequeuing respect passed in maxResources
* Ported over the new run props: idempotencyKeyExpiresAt, versions, oneTimeUseToken, maxDurationInSeconds
* Didn’t hit save… the new props when triggering tasks passed through
* Idempotency expiration + waitpoint edge case
* WIP on creating checkpoint, parking for now
* fix worker routes
* upgrade webapp node types to support generic event emitter
* separate event bus handler singleton and run failure alerts
* duration waits
* fix execution snapshot debug spans
* task waits
* fix event bus types
* temporary fix for react hook run handle type
* disable run notifications for now
* convert any typecasts to expect errors to more easily fix later
* fix webapp types after node types upgrade
* updateEnvConcurrencyLimits across marqs and the runqueue
* Pass proper values into the run engine
* RunQueue settings and removed unused rebalancing workers
* Remove rebalancing prop
* Tidied more things up
* Update/remove queue limits for MARQS and RunQueue
* taskQueue/concurrencyLimit changes ported back into the RunEngine
* Reworked completing waitpoints to improve performance and reduce race conditions
* Improved test robustness
* Down to a single run lock only when a run is totally unblocked and ready to continue
* warm starts, worker notifications, wait fixes
* Fix for Run Engine poll interval env var
* Expect the waitpoint to be completed quickly
* If a run is locked then it’s too late to expire it
* Added VALKEY_ env vars and plugged them into the run engine
* Extracted and updated the guard queue function so it can be used when batching
* Added logging and universal concurrency changes to trigger task v1
* Added notes back in
* Bump @trigger.dev/worker to 3.3.7
* reportInvocationUsage for the runAttemptStarted event
* improve execution snapshot span debug span start times
* Unfriendly IDs
* update lockfile
* Created a shared determineEngineVersion function
* disable unfinished commands
* save new cli config to different location, misc fixes
* add basic engine version check via current deploy
* new run engine will default to node 22 runtime
* block some actions for projects on previous run engine
* fix worker group tests
* fix triggerAndWait test
* one typescript version to rule them all
* redlock type patch
* fix type issues caused by ts-reset
* improve cleanup scripts
* add missing socket.io dep
* fix run notification handler type
* fix worker group test again
* generate prisma client for e2e tests
* remove worker group tests for now
* prevent image pull rate limits during unit tests
* increase timeout for queue concurrency limit test
* generate prisma client for preview release
* same node types everywhere
* Updated engine readme, removed legacy system notes
* use default machine preset from platform package
* worker instances plural in schema
* disable pnpm update notifications
* return worker group details from connect call
* add workers admin route
* fix heartbeat route return type
* move deployment labels to core apps
* refactor run controller env schema
* Add firstAttemptStartedAt to TaskRun
* RunEngine 2.0 batch trigger support (#1581)
* Make it clear when BatchTriggerV2Service is used
* Copy of BatchTriggerV2Service
* WIP batch triggering
* Allow blocking a run with multiple waitpoints at once. Made it atomic
* Removed unused param
* New batch service
* Pass through the parentRunId and resumeParentOnCompletion
* Use the new batch service, and correct trigger task version
* Force V1 engine if using BatchTriggerV2Service, we’ve already done the check at this point
* Removed the $transaction and early exit if nothing changed
* Adedd a simple batch task to the hello world reference catalog
* Fix for batch waits not working
* Added parentRunId in a couple more places
* Removed waitForBatch log
* Added another parentRunId
* Expanded the example to include all the different triggers
* More changes to blocking to support continuing after idempotent completed runs
* Fix for the wrong type when blocking a run
* remove @map
* optimise worker auth query
* add engine version header to core api client requests
* remove unique constraint for default group id
* consolidate migrations
* the first managed worker becomes the global default
* Debug events off by default, added an admin toggle to show them
* worker group name can't be an empty string
* add exec helper to core
* move machine resources to core
* add pre-dequeue callback to determine max resources
* optionally skip dequeue
* bump worker package
* move worker to core
* fix ReadableStream type error
* fix another type issue
* update a few more tsconfigs
* add metadata changes introduced in #1563
* Run Engine 2.0 trigger idempotency (#1613)
* Return isCached from the trigger API endpoint
* Fix for the wrong type when blocking a run
* Render the idempotent run in the inspector
* Event repository for idempotency
* Debug events off by default, added an admin toggle to show them
* triggerAndWait idempotency span
* Some improvements to the reference idempotency task
* Removed the cached tracing from the SDK
* Server-side creating cached span
* Improved idempotency test task
* Create cached task spans in a better way
* Idempotency span support inc batch trigger
* Simplified how the spans are done, using more of the existing code
* Improved the idempotency test task
* Added Waitpoint Batch type, add to TaskRunWaitpoint with order
* Pass batch ids through to the run engine when triggering
* Added batchIndex
* Better batch support in the run engine
* Added settings to batch trigger service, before major overhaul
* Allow the longer run/batch ids in the filters
* Changed how batching works, includes breaking changes in CLI
* Removed batch idempotency because it gets put on the runs instead
* Added `runs` to the batch.retrieve call/API
* Set firstAttemptStartedAt when creating the first attempt
* Do nothing when receiving a BATCH waitpoint
* Some fixes in the new batch trigger service… mostly just passing missing optional params through
* Tweaked the idempotency test task for more situations
* Only block with a batch if it’s a batchTriggerAndWait… 🤦♂️
* Added another case to the idempotency test task: multiple of the same idempotencyKey in a single batch
* Support for the same run multiple times in the same batch
* Small tweaks
* Make sure to complete batches, even if they’re not andWait ones
* Export RunDuplicateIdempotencyKeyError from the run engine
* Latest lockfile
* Trigger with a machine (old run engine)
* RE2, allow setting machine when triggering
* Fix for new glob patterns
* add max run count to dequeue from version route
* add worker instance name env var and header
* queue consumer pre skip callback
* poll for more runs after final execution errors
* fix dequeue search param schema
* add shortcut to debug switch
* expose run engine timeouts as env vars
* make warm start durations configurable
* add optional status to json reply helper
* fix preSkip hook, add debug logs
* BLOCKED_BY_WAITPOINTS -> SUSPENDED
* exit controller when run suspended
* check if already replied before http reply
* run controller will wait for next run after the current one is suspended
* cancel run button shortcut
* minimal event repository environment type
* fix update metadata call
* run suspension and misc fixes wip
* change debug shortcut to shift + D
* Started work on the Dev supervisor
* Formatting
* Fix for bad imports
* Before rebuilding SSE
* Presence updating from the CLI working via SSE
* add worker notification debug logs
* send run:stop when exiting run phase
* skip current snapshot poll on worker notification
* add more logs and route to submit run debug logs
* add worker and runner ids to snapshots
* improve run notification debug logs
* add workload debug log route
* misc run controller fixes and refactor
* prevent parallel execution of critical functions
* update bun to 1.2.1
* WIP with dev dequeuing
* Method to convert friendlyIds to non-friendly, do nothing with actual ids
* Set the engine on BackgroundWorker, lazily upgrade projects to engine V2
* Runs with ttls were getting immediately expired… oops.
* Pass the Waiting for deploy reason through, so we have it on the execution snapshots
* Fixed the logic for getting the right background worker for a run
* Use the correct ID when dequeuing…
* determineEngineVersion is now fully functional
* Rate limiter ignores the dev endpoints
* Retrieving a batch gives you the runIds
* Set a unique version for the RE2 BatchTaskRun
* add provisional changeset
* The start of dev run execution is working
* First dev run working
* Moved the dev run controller closer to what Nick did with the managed one
* export exec output type
* Heartbeat fix: don’t heartbeat if _isHeartbeating == false
* Dev runs get notifications, some dev bug fixes
* Improved logging or dequeuing
* We need to dequeue runs from the latest version too, for triggerAndWait
* Ported Eric’s validateWorkerManifest with nicer errors
* When flattening an idempotency key if part is undefined, return undefined
* Dev logging fixes
* Remove sigterm listener
* Deprecating workers. Don’t specify a BackgroundWorker when dequeuing an environment
* Deleted some old files. Renamed “managed” to “deploy”
* When a build finishes, always copy the build dir (otherwise the first one gets trampled on by the 2nd)
* Dev master queues should work differently
* Deleting old workers
* Added debounce function to core
* Improvement to canceling
* WIP on debounce canceling on socket disconnection
* Added environment data to execution snapshots
* Dev runs that have stalled get “Canceled” with a reason explaining why
* Show CLI messaged when a connection to the platform is lost/restored
* Fix TriggerTask after merge
* Add trigger task v2 max attempts, replace some findUniques
* Port the new queue logic to the run engine
* More fixes post-merge
* We weren’t setting a `retryConfig` up for the tests… it’s now required
* Start the Redis worker inside the Run Engine… 🤦♂️
* Trying to make the testcontainers more reliable
* Added keyPrefix: "engine:”
* Badly placed bracket in trigger task
* Better Redis namespacing
* Fix for expired run not getting removed from the queue
* Don’t create a redis client in the testcontainers, return the redisOptions instead
* Cleanup redis client in the run lock tests
* Fix for the RunQueue not supporting keyPrefix
* Updated more of the RunQueue scripts rebalancing
* Trying to make Redis more robust in the tests…
* Improved test resiliciency more
* Fix for delays (checkpoint check)
* Increase the timeout slightly to fix ttl test
* Added priority support when triggering
* More wip trying to make test containers more reliable
* batchTriggerAndWait test is still failing… some wip to try fix it
* Fixed redis tests now we’re not providing a client
* Separate Redis clients for the run engine worker/queue/runlock
* Made the wait for duration test more resilient
* Added idempotencyKeyExpiresAt to Waitpoints
* Waitpoint timeouts and idempotency expiry
* Use finishWaitpoint, removed extra worker job
* Added waitpoint idempotency tests
* Creating resume tokens is working
* Some improvements to the resume tokens
* Moved resumeTokens to just be wait functions 🥳
* Delete old RuntimeManagers
* Wait for token is working
* Better test for the wait tokens
* Improved the test task some more
* Hide the accessories in the span inspector
* WIP on waitpoint inspector
* WIP on complete waitpoint form
* Span overview panel can be changed based on the entity type
* Improved the waitpoint display
* WIP on completing waitpoint form
* Use the existing CodeBlock for the tip
* Style improvements
* Complete waitpoint
* All waitpoint sidebar variants
* Waits now use a pause icon
* Durations waits use the API to create/block with a waitpoint, not the runtime
* Fix for engine.blockRunWithWaitpoint required org id
* Removed old wait code from the run controllers/task run process
* Form action for skipping a datetime waitpoint
* Move testDockerCheckpoint to a separate core package export (it can’t be bundled on the client)
* Fix for glitchy hourglass animation
* Completed waitpoints display better
* Increase Redis maxRetriesPerRequest to 20 (default)
* Completing and skipping waitpoints is working
* Remove the database prisma dev command, since we need to use create only now. Updated docs
* Added skip timeout, reworked the UI
* Tweaked spacing
* Added payload limit to waitpoint token completion from dashboard
* Test idempotency works on wait.for and wait.until
* Moved the worker-actions to /engine/ from /api/
* Moved dev engine endpoints to /engine/ from /api/
* Separate /engine/ rate limiter
* Added parallel wait prevention, it’s working for duration waits but not well for triggerAndWait yet
* WIP post-merge conflicts
* Set taskEventStore column in the new engine
* Remove duplicate keys
* Post-merge fixes
* Fix for span merge layout
* Use executedAt instead of firstAttemptStartedAt
---------
Co-authored-by: Matt Aitken <matt@mattaitken.com>
* Fix streaming splits in realtime streams v2
* Add changeset
* Skip all flaky tests 😡
* Improve the way we stream from tasks to the server
* Improve the v1 realtime streams (Redis)
* Turn on the relay realtime stream service
* Improved the relay realtime cleanup
* Fixed consuming realtime runs w/streams after the run is already finished
* Remove some logs
* Update changeset
* Fixed runStream tests
* WIP realtime streams
* Handle realtime with large payloads or outputs #1451
* feat: optimize Redis stream handling with batching
Add STREAM_ORIGIN to environment schema. Improve performance in
RealtimeStreams by using TextDecoderStream for simpler text
decoding and implementing batching of XADD commands for Redis
streams. Limit stream size using MAXLEN option. Update
environment variable repository with new variable type. Adjust
import statements for Redis key and value types.
* 🔧 chore: add dev dependencies for bundle analysis
* add metadata tests and a few more utilties
* Add stream tests and improve streaming
* Added AI tool tasks, descriptions to tasks
* Use the config file path to determine the workingDir, then the package.json path
* Remove stream test files
* useTaskTrigger react hook that allows triggering a task from the client
* Add streaming support for the realtime react hooks
* Add ability to stream results after useTaskTrigger
* Improve the stream throttling
* Use the runId as the ID key to bust the cache after triggering
* Upgrade to to the latest electric sql client and server
* Make realtime server backwards compat with 3.1.2 release
* Pass the runId into useRealtimeRun
* Fix scopes when specifiying reading all runs
* WIP @trigger.dev/rsc package
* Various fixes and accepted recommendations by CodeRabbit
* Regenerate pnpm lock file
* A couple tweaks to rsc and give up on rendering react in tasks for now
* Add changeset
* Remove triggerRequest from the useEffect deps
* Improve realtime & frontend authentication errors
* Fixed authorization tests
* Remove unnecessary log
* Add metadata.stream limits and improve the metadata streams structure
* Streams can now have up to 2500 entries
* Various coderabbit fixes
* additional react-hooks jsdocs
* Denormalize run tags, increase character limit to 128
* WIP realtime subscribing to runs
* extracted the stream stuff into core, made it more reusable
* WIP tags
* Remove tags for now because it’s not support in electric
* Support async iterables, readable stream, and callback style subscription styles
* Remove tags streaming endpoint
* Add realtime rate limits and scope them to the /realtime path
* WIP rate limt per org
* Introduce per org rate limits
* WIP JWT auth
* Move migrations into new internal db package
* Resolve pnpm lock file
* Authenticating to the realtime API with JWTs are working
* realtime in the client
* Created react-hooks package and starting to move stuff in there
* Improve types for hooks
* schema tasks
* Added useBatch hook
* build uploadthing/fal demo and change how run metadata is synced to the server
* tweaks
* WIL realtime concurrency tracking
* Implement test for realtime client using testcontainers
also updated electric to latest version
* Allow customizing the expiration time of the automatic JWT created after triggering a task
* Add support for subscribing to run tags
* Improve auth types and API
* finalize the realtime API
* Fixed some example stuff
* Allow up to 10 run tags
* Remove core from docker-provider tsconfig paths to prevent it from being typechecked
* do the same for the kubernetes provider
* Fixing some typecheck errors
* Fix webapp type errors
* Update @trigger.dev/platform to 1.0.13
* Fix attw error
* Remove from/to in subscribeToRuns query params
* Add tests for the rate limit middleware and add custom JWT rate limits
* turn off webapp test parallelism
* Finish renaming jwt -> publicAccessToken and automatically give the JWT read access to the tags when using trigger
* Add changeset
* Attempt to fix unit tests in CI
* Skip running the auth rate limit middleware tests for now
* Try a beefier machine
* Try and run webapp tests separately
* Setup env vars
* Make sliding window test more reliabile
* v3: cancel subtasks when parent task runs are cancelled
* v3: recover from server rate limiting errors in a more reliable way
- Changing from sliding window to token bucket in the API rate limiter, to help smooth out traffic
- Adding spans to the API Client core & SDK functions
- Added waiting spans when retrying in the API Client
- Retrying in the API Client now respects the x-ratelimit-reset
- Retrying ApiError’s in tasks now respects the x-ratelimit-reset
- Added AbortTaskRunError that when thrown will stop retries
- Added idempotency keys SDK functions and automatically injecting the run ID when inside a task
- Added the ability to configure ApiRequestOptions (retries only for now) globally and on specific calls
- Implement the maxAttempts TaskRunOption (it wasn’t doing anything before)
* Adding some docs about the request options
* Fix type error
* Remove context propagation through graphile jobs
* Remove logger
* only select a subset of task run columns
* limit columns selected in batchTrigger as well
* added idempotency doc
* allow scoped idempotency keys, and fixed an issue with the unique index on BatchTaskRun and TaskRun
* Removed old cancel task run children code
* Trying to use the @upstash/ratelimit package with ioredis…
* WIP using the redis package instead
* Revert the action back
* Removed redis
* Started refactoring
* SQS setting for the poll interval. Set the default queue reading to be slower
* API rate limiter as Express middleware
* Organise imports
* Fixed spelling mistake “limitter”
* No authorization header response is problem+json
* WIP proxy/deploy
* WIP, registry proxy in express working
* A couple of notes, preparing for indexing
* Move the changes to prod-worker into the new file
* Deploy command working with indexing and runs (docker provider only for now)
* Removed ts-expect-error directive
* Fixed build command
* Downgrade to remix-auth-email-link to remove yarn dependency
* Turn off the pg listen service for now
* Add snapshot admin route
* A couple logger fixes
* Add ability to disable compression
* Add ability to disable SSE
* Fix SSE memory leak + DB load issue
* Implement manually invokable jobs through the invokeTrigger
Also implemented a job run notification system, that will POST details of a run on completion. This combines with the task callbackUrl system to implement the invokeAndWait
* Document the invoke trigger
* batch invoke and wait
* background fetch timeouts
* Use @whatwg-node/fetch instead of the polyfilled fetch
* Fix some outdated dependencies in webapp
* Improved subtask error propogation messages
* Document the OpenAI changes and the batch invoke stuff
* Fix dequeuing jobs
* Don’t retry the OpenAI completion background task
* Added OpenAI changesets
* Use the new ResumeTaskService in ProcessCallbackTimeout as well
* Auto-yield run execution to help prevent duplicate task executions
* Add auto-yield config to endpoints
* Refactor run execution with buffer and limits
Introduced constants RUN_CHUNK_EXECUTION_BUFFER and MAX_RUN_CHUNK_EXECUTION_LIMIT. Adjusted PerformRunExecutionV2Service to use the new constants to fine-tune execution timings and buffers.
* Add endpoint probing functionality
Added new `RESPONSE_TIMEOUT_STATUS_CODES` in `consts.ts` to manage timeout responses. Additional functions `detectResponseIsTimeout(response: Response)` was added in `endpoint.server.ts` to detect if a response was a timeout based on the status codes from `RESPONSE_TIMEOUT_STATUS_CODES`.
Update actions to use new endpoint probing endpoint service. This allows for the early probing of endpoints to determine if they're up and running.
A new class `ProbeEndpointService` was created in `probeEndpoint.server.ts` which makes HTTP requests to a given endpoint and updates its properties based on the result.
Finally, `detectResponseIsTimeout(response)` is used in `performRunExecutionV2.server.ts` for marking the execution as succeeded when facing a timeout.
* Refactored probe method in EndpointApi class
The probe method of the EndpointApi class has been refactored to remove the error handling part and it now takes a timeout sent from the client directly. The corresponding changes were also made in the ProbeEndpointService and TriggerClient objects to reflect the alterations in the probe method.
The error handling related to the timeout has been removed and the responsibility of handling the timeout has been shifted to the client. Thus, the probe method has been greatly simplified. The `probeEndpoint.server.ts` file was also changed to accommodate the change in behavior of the probe result.
In the `triggerClient.ts` the timeout for probe is now read from the incoming request object. For backward compatibility, if no timeout is provided in the request, the default value of 15 minutes is used.
* Remove performRunExecution v1 enqueue function
* Better document limits and add docs on increasing function timeouts
* Upgrade webapp docker container to use 18.18.2
* force clients to yield when a run is executing in a gracefully shutting down worker
* Renamed task `key` to `cacheKey` and added more task documentation
* Index the `@trigger.dev/sdk` version on Endpoints
* WIP
* Report ECS task info on startup and shutdown
* Fixed lifecycle name
* Re-add terminus
* Require the build dir when http server is disabled
* Remove unnecessary logs
* Implement graceful shutdown in ZodWorker
* Re-order some code
* Increase the keepAliveTimeout to 65 seconds to prevent LB 502 errors
* WIP job run performance improvements
- Added a `perf` tool to better measure job run performance under heavy load
- Removed `runFinished` job (not really needed)
- startQueuedRuns now uses a jobKey with replace
- Fixed an issue with ZodWorker when using jobKey
* Publish improvement docker images
* fixed the improvement docker publishing
* Downgrade back to prisma 4.16.0 because 5.1.x broke docker builds
* Changes to how queued runs work
- Split the worker into two different workers, one dedicated to performRunExecution
- Schedule performRunExecution in a single place, with a queue and using a round robin manually controlled concurrency
- Remove startQueuedRuns
- All runs are queued before they are started
- Setting the worker maxPoolSize to the same as the worker concurrency
- Starting to be able to split the docker image
* Remove queue name from startRun graphile job
* Make the prisma connection pool stuff configurable through env vars
* Hardcode (for now) the max concurrent runs limit
* Rewrite performRunExecution to be more performant
PerformRunExecutionV2:
- Does not create and manage jobRunExecution records
- Does not reimplement retrying, uses graphile worker retrying instead
I’ve kept around PerformRunExecutionV1 so this works when deploying. Definitely needs LOTS of testing
* Fix issues with cached tasks
- Limit the size of the cached tasks sent when executing a run, using the knapsack problem dynamic programming approach
- Actually USE the cached tasks in IO by using the idempotencyKey instead of the task ID
- Remove output from all logs
- Added a stress test job catalog
* Forgot to commit the logger updates
* Never log connectionString
* Login to docker hub to get around rate limits
* Add additional logging to the graphile workers
* Fix the *_ENABLED env vars
* Allow adding and removing jobs to be done from the webapp
* Don’t set the job to failed if it’s being retried
* Deprecated queue options in the job and removed startPosition. Now using the job/env combo as the job queue name
* Dequeung jobs doesn’t check if the runner is initialized
* Fixed issues with retrying a run getting stuck on a cancelled task, and errors from parsing the results of dequeing a job
* Remove queued round robin thing that isn’t used anymore
* Added slack to job catalog
* Better forwards compat
* Added long delay
* Fixed lock file
* Setup project-wide prettier
* Remove old workspace file
* Remove old debugging directives
* New top-level .prettierignore
* Updated Prettier config settings
* Contrubuting guide: Fix for some bad code blocks
* Added more ignores
* Improved the format script command
* printWidth set to 100
* Formatted entire repo (pnpm run format)
Chris Arderne