chore: release v4.5.0 (#3998)

# Trigger.dev v4.5.0

4.5.0 is the GA of the AI Agents platform. Everything built during the
prerelease line (durable agents, Sessions, AI Prompts) is now stable on
the `latest` tag, alongside a set of SDK and runtime improvements.

## AI Agents (`chat.agent`)

Run Vercel AI SDK chat completions as durable Trigger.dev tasks instead
of fragile API routes. A conversation runs as one long-lived task keyed
on `chatId`, so it survives page refreshes, network blips, redeploys,
and crashes, and every turn is a span in the dashboard.

```ts
import { chat } from "@trigger.dev/sdk/ai";
import { streamText, stepCountIs } from "ai";
import { anthropic } from "@ai-sdk/anthropic";

export const myChat = chat.agent({
  id: "my-chat",
  run: async ({ messages, signal }) => {
    return streamText({
      ...chat.toStreamTextOptions(), // system prompt, compaction, steering, telemetry
      model: anthropic("claude-sonnet-4-5"),
      messages,
      abortSignal: signal,
      stopWhen: stepCountIs(15),
    });
  },
});
```

## Sessions

The durable primitive underneath `chat.agent`, usable on its own: a
run-aware, bidirectional stream channel keyed on a stable `externalId`
whose `.in` / `.out` streams survive run boundaries (suspend, crash,
idle-timeout, redeploy). One Session spans many runs, which makes it a
good fit for agent inboxes and approval flows.

```ts
import { sessions } from "@trigger.dev/sdk";

// Create the session and trigger its first run (idempotent on externalId)
await sessions.start({
  type: "inbox",
  externalId: userId,
  taskIdentifier: "inbox-agent",
});

const session = sessions.open(userId);
await session.in.send({ text: "hello" });

const stream = await session.out.read({ signal: AbortSignal.timeout(30_000) });
for await (const chunk of stream) console.log(chunk); // durable across run swaps
```

## AI Prompts

Define prompt templates as code, versioned on every deploy, and override
the text or model from the dashboard without redeploying
(environment-scoped). Each generation links back to its prompt version
for usage, cost, and latency.

```ts
import { prompts } from "@trigger.dev/sdk";
import { z } from "zod";

export const supportPrompt = prompts.define({
  id: "customer-support",
  model: "gpt-4o",
  variables: z.object({ customerName: z.string(), issue: z.string() }),
  content: `You are a support agent for Acme.
Customer: {{customerName}}
Issue: {{issue}}`,
});

// Honors any active dashboard override, else the current deployed version
const resolved = await supportPrompt.resolve({ customerName: "Alice", issue: "Can't log in" });
// resolved.text, resolved.model, resolved.version
```

## `useChat` integration

`useTriggerChatTransport` is a Vercel AI SDK `ChatTransport` that runs
`useChat` over Trigger.dev realtime with no API routes. Text, tool
calls, reasoning, and `data-*` parts stream natively, and it works with
AI SDK v5, v6, and now v7.

## First-turn fast path (`chat.headStart`)

Runs the first turn in your warm server process while the agent boots in
parallel, cutting cold-start time-to-first-chunk roughly in half
(measured ~2.8s to ~1.2s). Available via the new
`@trigger.dev/sdk/chat-server` subpath.

## Human-in-the-loop, stop, and steering

The agent control surface: tool approvals (`needsApproval` +
`addToolApprovalResponse`), client-driven stop-generation, mid-execution
steering (`pendingMessages`), and between-turn context injection
(`chat.inject` / `chat.defer`), all durable across the conversation.

## Agent Skills

`skills.define({ id, path })` bundles a `SKILL.md` folder into your
deploy image. The agent gets a one-line summary up front and loads the
full instructions plus scoped `bash` / `readFile` tools on demand
(progressive disclosure), so a capability is something the model reaches
for rather than a pre-declared typed tool.

## `trigger skills` for coding assistants

`trigger skills` installs version-pinned Trigger.dev skills plus a
bundled docs snapshot into Claude Code, Cursor, GitHub Copilot, and
Codex, so your assistant's Trigger.dev knowledge stays current with your
installed SDK version. `trigger init` now offers to set up the MCP
server and skills too.

## Model library

A new Models page in the dashboard: a catalog of models grouped by
provider with context window, capabilities, and input / output pricing
per 1M tokens, plus a "Your models" tab showing per-model usage, cost,
and cache-hit sparklines from your actual traffic.

## Dev branches

Run multiple local `trigger dev` sessions in parallel (separate git
worktrees or coding agents) without runs colliding, each isolated with
its own dashboard, via `trigger dev --branch <name>`.

## `TriggerClient`

An instantiable client so one process can trigger and read across
projects, environments, and preview branches, each with its own auth and
baseURL, with no shared global state.

```ts
import { TriggerClient } from "@trigger.dev/sdk";

const prod = new TriggerClient({ accessToken: process.env.TRIGGER_PROD_KEY });
const preview = new TriggerClient({
  accessToken: process.env.TRIGGER_PREVIEW_KEY,
  previewBranch: "signup-flow",
});

await prod.tasks.trigger("send-email", { to: "user@example.com" });
await preview.runs.list({ status: ["COMPLETED"] });
```

## SDK and runtime

- AI SDK 7 support (v5 and v6 still supported), with OpenTelemetry
telemetry auto-wired
- Large trigger-payload offload: trigger payloads at or above 128KB
upload to object storage automatically, using the same auth and baseURL
as the trigger call
- Region support on the runs API: filter runs by region and read each
run's executing region (also on MCP `list_runs`)
- Duplicate task-id detection: `dev` and `deploy` fail with a clear
error instead of silently overwriting
- `envvars.upload` gains an `isSecret` flag to import redacted secret
variables
- Retry hardening: `TASK_MIDDLEWARE_ERROR` now retries under the task's
retry policy

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

.changeset/agent-skills-bundled-in-sdk.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"trigger.dev": patch
----
-
-`@trigger.dev/sdk` now bundles the Trigger.dev agent skills and a curated snapshot of the docs those skills reference. The skills that `trigger skills` installs into your coding agent read this content from node_modules, so the guidance your AI assistant follows is pinned to the SDK version installed in your project and stays current across upgrades instead of going stale until the next reinstall.

.changeset/agent-skills.md

@@ -1,16 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"@trigger.dev/core": patch
-"@trigger.dev/build": patch
-"trigger.dev": patch
----
-
-Add Agent Skills for `chat.agent`. Drop a folder with a `SKILL.md` and any helper scripts/references next to your task code, register it with `skills.define({ id, path })`, and the CLI bundles it into the deploy image automatically — no `trigger.config.ts` changes. The agent gets a one-line summary in its system prompt and discovers full instructions on demand via `loadSkill`, with `bash` and `readFile` tools scoped per-skill (path-traversal guards, output caps, abort-signal propagation).
-
-```ts
-const pdfSkill = skills.define({ id: "pdf-extract", path: "./skills/pdf-extract" });
-
-chat.skills.set([await pdfSkill.local()]);
-```
-
-Built on the [AI SDK cookbook pattern](https://ai-sdk.dev/cookbook/guides/agent-skills) — portable across providers. SDK + CLI only for now; dashboard-editable `SKILL.md` text is on the roadmap.

.changeset/ai-prompts.md

@@ -1,52 +0,0 @@
----
-"@trigger.dev/sdk": minor
----
-
-**AI Prompts** — define prompt templates as code alongside your tasks, version them on deploy, and override the text or model from the dashboard without redeploying. Prompts integrate with the Vercel AI SDK via `toAISDKTelemetry()` (links every generation span back to the prompt) and with `chat.agent` via `chat.prompt.set()` + `chat.toStreamTextOptions()`.
-
-```ts
-import { prompts } from "@trigger.dev/sdk";
-import { generateText } from "ai";
-import { openai } from "@ai-sdk/openai";
-import { z } from "zod";
-
-export const supportPrompt = prompts.define({
-  id: "customer-support",
-  model: "gpt-4o",
-  config: { temperature: 0.7 },
-  variables: z.object({
-    customerName: z.string(),
-    plan: z.string(),
-    issue: z.string(),
-  }),
-  content: `You are a support agent for Acme.
-
-Customer: {{customerName}} ({{plan}} plan)
-Issue: {{issue}}`,
-});
-
-const resolved = await supportPrompt.resolve({
-  customerName: "Alice",
-  plan: "Pro",
-  issue: "Can't access billing",
-});
-
-const result = await generateText({
-  model: openai(resolved.model ?? "gpt-4o"),
-  system: resolved.text,
-  prompt: "Can't access billing",
-  ...resolved.toAISDKTelemetry(),
-});
-```
-
-**What you get:**
-
-- **Code-defined, deploy-versioned templates** — define with `prompts.define({ id, model, config, variables, content })`. Every deploy creates a new version visible in the dashboard. Mustache-style placeholders (`{{var}}`, `{{#cond}}...{{/cond}}`) with Zod / ArkType / Valibot-typed variables.
-- **Dashboard overrides** — change a prompt's text or model from the dashboard without redeploying. Overrides take priority over the deployed "current" version and are environment-scoped (dev / staging / production independent).
-- **Resolve API** — `prompt.resolve(vars, { version?, label? })` returns the compiled `text`, resolved `model`, `version`, and labels. Standalone `prompts.resolve<typeof handle>(slug, vars)` for cross-file resolution with full type inference on slug and variable shape.
-- **AI SDK integration** — spread `resolved.toAISDKTelemetry({ ...extra })` into any `generateText` / `streamText` call and every generation span links to the prompt in the dashboard alongside its input variables, model, tokens, and cost.
-- **`chat.agent` integration** — `chat.prompt.set(resolved)` stores the resolved prompt run-scoped; `chat.toStreamTextOptions({ registry })` pulls `system`, `model` (resolved via the AI SDK provider registry), `temperature` / `maxTokens` / etc., and telemetry into a single spread for `streamText`.
-- **Management SDK** — `prompts.list()`, `prompts.versions(slug)`, `prompts.promote(slug, version)`, `prompts.createOverride(slug, body)`, `prompts.updateOverride(slug, body)`, `prompts.removeOverride(slug)`, `prompts.reactivateOverride(slug, version)`.
-- **Dashboard** — prompts list with per-prompt usage sparklines; per-prompt detail with Template / Details / Versions / Generations / Metrics tabs. AI generation spans get a custom inspector showing the linked prompt's metadata, input variables, and template content alongside model, tokens, cost, and the message thread.
-
-See [/docs/ai/prompts](https://trigger.dev/docs/ai/prompts) for the full reference — template syntax, version resolution order, override workflow, and type utilities (`PromptHandle`, `PromptIdentifier`, `PromptVariables`).

.changeset/ai-sdk-7-support.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Adds AI SDK 7 support. The `ai` peer range now includes v7, and the `chat.agent` / chat surfaces work against v7's ESM-only build. On v7, install `@ai-sdk/otel` alongside `ai` and the SDK registers it for you so `experimental_telemetry` spans keep flowing into your run traces (v7 stopped emitting them from `ai` core). v5 and v6 keep working unchanged.

.changeset/ai-tool-helpers.md

@@ -1,15 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Add `ai.toolExecute(task)` so you can wire a Trigger subtask in as the `execute` handler of an AI SDK `tool()` while defining `description` and `inputSchema` yourself — useful when you want full control over the tool surface and just need Trigger's subtask machinery for the body.
-
-```ts
-const myTool = tool({
-  description: "...",
-  inputSchema: z.object({ ... }),
-  execute: ai.toolExecute(mySubtask),
-});
-```
-
-`ai.tool(task)` (`toolFromTask`) keeps doing the all-in-one wrap and now aligns its return type with AI SDK's `ToolSet`. Minimum `ai` peer raised to `^6.0.116` to avoid cross-version `ToolSet` mismatches in monorepos.

.changeset/backpressure-scale-up-freeze.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Add optional `shouldPauseScaling` to the supervisor consumer pool scaling options to freeze scale-up while it returns true (scale-down stays allowed).

.changeset/bundle-skills-single-pass.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-Fix `chat.agent` skills silently missing in `trigger dev` for projects whose task files read `process.env` at module top level (e.g. a third-party SDK client initialized at import). Skill folders now bundle into `.trigger/skills/` reliably regardless of which env vars are set when the CLI launches.

.changeset/cap-idempotency-key-length.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Reject overlong `idempotencyKey` values at the API boundary so they no longer trip an internal size limit on the underlying unique index and surface as a generic 500. Inputs are capped at 2048 characters — well above what `idempotencyKeys.create()` produces (a 64-character hash) and above any realistic raw key. Applies to `tasks.trigger`, `tasks.batchTrigger`, `batch.create` (Phase 1 streaming batches), `wait.createToken`, `wait.forDuration`, and the input/session stream waitpoint endpoints. Over-limit requests now return a structured 400 instead.

.changeset/chat-agent-hardening.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"@trigger.dev/core": patch
----
-
-Reliability fixes for `chat.agent`. A user message sent while the agent is streaming is no longer delivered twice (which could run a duplicate turn), input appends now carry an idempotency key so a retried send can't duplicate a message, stopping a generation clears the streaming state so a page reload doesn't replay the stopped turn, and runs can now carry the full set of dashboard tags instead of being silently truncated. `onTurnComplete` now fires on errored turns (with the thrown error attached) and the failed turn's user message is persisted so it isn't lost on the next run. Custom agents and manual `chat.writeTurnComplete` callers now trim the output stream, sending a custom action no longer leaves a second stream reader running, and a long-lived `watch` subscription no longer grows its dedupe set without bound.

.changeset/chat-agent-on-boot-hook.md

@@ -1,21 +0,0 @@
----
-"@trigger.dev/sdk": minor
----
-
-Adds `onBoot` to `chat.agent` — a lifecycle hook that fires once per worker process picking up the chat. Runs for the initial run, preloaded runs, AND reactive continuation runs (post-cancel, crash, `endRun`, `requestUpgrade`, OOM retry), before any other hook. Use it to initialize `chat.local`, open per-process resources, or re-hydrate state from your DB on continuation — anywhere the SAME run picking up after suspend/resume isn't enough.
-
-```ts
-const userContext = chat.local<{ name: string; plan: string }>({ id: "userContext" });
-
-export const myChat = chat.agent({
-  id: "my-chat",
-  onBoot: async ({ clientData, continuation }) => {
-    const user = await db.user.findUnique({ where: { id: clientData.userId } });
-    userContext.init({ name: user.name, plan: user.plan });
-  },
-  run: async ({ messages, signal }) =>
-    streamText({ model: openai("gpt-4o"), messages, abortSignal: signal }),
-});
-```
-
-Use `onBoot` (not `onChatStart`) for state setup that must run every time a worker picks up the chat — `onChatStart` fires once per chat and won't run on continuation, leaving `chat.local` uninitialized when `run()` tries to use it.

.changeset/chat-agent-preview-branch.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Fix `chat.agent` / `AgentChat` when the agent is deployed to a Trigger.dev preview branch. The realtime message-append and stream-subscribe calls now send the `x-trigger-branch` header (sourced from the same resolver `sessions.start` uses), so messaging a preview-branch chat agent no longer fails with `x-trigger-branch header required for preview env`.

.changeset/chat-agent-tools.md

@@ -1,15 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Add a `tools` option to `chat.agent`. Declaring your tools here threads them into the SDK's internal `convertToModelMessages`, so each tool's `toModelOutput` is re-applied when prior-turn history is re-converted.
-
-```ts
-chat.agent({
-  tools: { readFile, search },
-  run: async ({ messages, tools, signal }) =>
-    streamText({ model, messages, tools, abortSignal: signal }),
-});
-```
-
-Also exports `InferChatUIMessageFromTools<typeof tools>` to derive the chat `UIMessage` type (typed tool parts) directly from a tool set.

.changeset/chat-agent.md

@@ -1,44 +0,0 @@
----
-"@trigger.dev/sdk": minor
-"@trigger.dev/core": patch
----
-
-**AI Agents** — run AI SDK chat completions as durable Trigger.dev agents instead of fragile API routes. Define an agent in one function, point `useChat` at it from React, and the conversation survives page refreshes, network blips, and process restarts.
-
-```ts
-import { chat } from "@trigger.dev/sdk/ai";
-import { streamText } from "ai";
-import { openai } from "@ai-sdk/openai";
-
-export const myChat = chat.agent({
-  id: "my-chat",
-  run: async ({ messages, signal }) =>
-    streamText({ model: openai("gpt-4o"), messages, abortSignal: signal }),
-});
-```
-
-```tsx
-import { useChat } from "@ai-sdk/react";
-import { useTriggerChatTransport } from "@trigger.dev/sdk/chat/react";
-
-const transport = useTriggerChatTransport({ task: "my-chat", accessToken, startSession });
-const { messages, sendMessage } = useChat({ transport });
-```
-
-**What you get:**
-
-- **AI SDK `useChat` integration** — a custom [`ChatTransport`](https://sdk.vercel.ai/docs/ai-sdk-ui/transport) (`useTriggerChatTransport`) plugs straight into Vercel AI SDK's `useChat` hook. Text streaming, tool calls, reasoning, and `data-*` parts all work natively over Trigger.dev's realtime streams. No custom API routes needed.
-- **First-turn fast path (`chat.headStart`)** — opt-in handler that runs the first turn's `streamText` step in your warm server process while the agent run boots in parallel, cutting cold-start TTFC by roughly half (measured 2801ms → 1218ms on `claude-sonnet-4-6`). The agent owns step 2+ (tool execution, persistence, hooks) so heavy deps stay where they belong. Web Fetch handler works natively in Next.js, Hono, SvelteKit, Remix, Workers, etc.; bridge to Express/Fastify/Koa via `chat.toNodeListener`. New `@trigger.dev/sdk/chat-server` subpath.
-- **Multi-turn durability via Sessions** — every chat is backed by a durable Session that outlives any individual run. Conversations resume across page refreshes, idle timeout, crashes, and deploys; `resume: true` reconnects via `lastEventId` so clients only see new chunks. `sessions.list` enumerates chats for inbox-style UIs.
-- **Auto-accumulated history, delta-only wire** — the backend accumulates the full conversation across turns; clients only ship the new message each turn. Long chats never hit the 512 KiB body cap. Register `hydrateMessages` to be the source of truth yourself.
-- **Lifecycle hooks** — `onPreload`, `onChatStart`, `onValidateMessages`, `hydrateMessages`, `onTurnStart`, `onBeforeTurnComplete`, `onTurnComplete`, `onChatSuspend`, `onChatResume` — for persistence, validation, and post-turn work.
-- **Stop generation** — client-driven `transport.stopGeneration(chatId)` aborts mid-stream; the run stays alive for the next message, partial response is captured, and aborted parts (stuck `partial-call` tools, in-progress reasoning) are auto-cleaned.
-- **Tool approvals (HITL)** — tools with `needsApproval: true` pause until the user approves or denies via `addToolApprovalResponse`. The runtime reconciles the updated assistant message by ID and continues `streamText`.
-- **Steering and background injection** — `pendingMessages` injects user messages between tool-call steps so users can steer the agent mid-execution; `chat.inject()` + `chat.defer()` adds context from background work (self-review, RAG, safety checks) between turns.
-- **Actions** — non-turn frontend commands (undo, rollback, regenerate, edit) sent via `transport.sendAction`. Fire `hydrateMessages` + `onAction` only — no turn hooks, no `run()`. `onAction` can return a `StreamTextResult` for a model response, or `void` for side-effect-only.
-- **Typed state primitives** — `chat.local<T>` for per-run state accessible from hooks, `run()`, tools, and subtasks (auto-serialized through `ai.toolExecute`); `chat.store` for typed shared data between agent and client; `chat.history` for reading and mutating the message chain; `clientDataSchema` for typed `clientData` in every hook.
-- **`chat.toStreamTextOptions()`** — one spread into `streamText` wires up versioned system [Prompts](https://trigger.dev/docs/ai/prompts), model resolution, telemetry metadata, compaction, steering, and background injection.
-- **Multi-tab coordination** — `multiTab: true` + `useMultiTabChat` prevents duplicate sends and syncs state across browser tabs via `BroadcastChannel`. Non-active tabs go read-only with live updates.
-- **Network resilience** — built-in indefinite retry with bounded backoff, reconnect on `online` / tab refocus / bfcache restore, `Last-Event-ID` mid-stream resume. No app code needed.
-
-See [/docs/ai-chat](https://trigger.dev/docs/ai-chat/overview) for the full surface — quick start, three backend approaches (`chat.agent`, `chat.createSession`, raw task), persistence and code-sandbox patterns, type-level guides, and API reference.

.changeset/chat-boot-cursor.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"@trigger.dev/core": patch
----
-
-Continuation chat boots no longer stall for around 10 seconds before the first turn. The `session.in` resume cursor is now found with a non-blocking records read instead of draining an SSE long-poll (which always waited out its full 5 second inactivity window, twice per boot), the boot reads run concurrently, and chat snapshots carry the cursor so subsequent boots skip the scan entirely.

.changeset/chat-head-start-prepare-messages.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Fix Head Start handovers breaking when a `chat.agent` also defines a `prepareMessages` hook. A handover hands the first turn's pending tool call to the agent as a tool-approval round whose trailing tool message must reach the model untouched. A `prepareMessages` hook that rewrites the last message (for example the recommended prompt-caching breakpoint) could disturb it, so the turn failed with "tool_use ids were found without tool_result". The agent now preserves that approval tail across `prepareMessages`, so caching and Head Start compose cleanly.

.changeset/chat-headstart-api-client.md

@@ -1,14 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-`chat.headStart` now accepts an `apiClient` option (base URL + access token), so the head-start route can create the session and trigger the agent run against a different project/environment than the warm server's ambient Trigger config. Useful when your `chat.agent` lives in a separate project from the app serving the route. Mirrors the `apiClient` option on `chat.createStartSessionAction`; your LLM provider keys stay in the `run` callback and are unaffected.
-
-```ts
-export const POST = chat.headStart({
-  agentId: "my-agent",
-  apiClient: { baseURL, accessToken },
-  run: async ({ chat }) =>
-    streamText({ ...chat.toStreamTextOptions({ tools }), model: anthropic("claude-sonnet-4-6") }),
-});
-```

.changeset/chat-headstart-custom-backends.md

@@ -1,26 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-`chat.headStart` now works with the `chat.customAgent` and `chat.createSession` backends, not only `chat.agent`. The warm step-1 response hands over to your loop the same way it does for a managed agent.
-
-In a `chat.customAgent` loop, consume the handover on turn 0:
-
-```ts
-const conversation = new chat.MessageAccumulator();
-const { isFinal, skipped } = await conversation.consumeHandover({ payload });
-if (skipped) return; // warm handler aborted, so exit without a turn
-if (isFinal) {
-  await chat.writeTurnComplete(); // step 1 is the response, no streamText
-} else {
-  const result = streamText({ model, messages: conversation.modelMessages, tools });
-  // Pass originalMessages so the handed-over tool round merges into the
-  // step-1 assistant instead of starting a new message.
-  const response = await chat.pipeAndCapture(result, {
-    originalMessages: conversation.uiMessages,
-  });
-  if (response) await conversation.addResponse(response);
-}
-```
-
-With `chat.createSession`, the iterator surfaces it as `turn.handover`; call `turn.complete()` with no argument on a final handover. The lower-level `chat.waitForHandover()` and `accumulator.applyHandover()` are also exported for hand-rolled loops.

.changeset/chat-headstart-hydrate.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Fix `chat.headStart` when `hydrateMessages` is registered. The warm route's step-1 partial now reaches the agent's accumulator on the hydrate path, so `onTurnComplete` carries the full first turn (the head-start user message included), tool-call handovers resume from step 2 instead of re-running step 1, and the assistant `messageId` stays stable across the handover.

.changeset/chat-headstart-reasoning.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Preserve reasoning parts across the `chat.headStart` handover. Extended-thinking models' step-1 reasoning now lands in the durable session history (and `onTurnComplete`) under the same assistant `messageId`, with provider metadata intact so Anthropic thinking signatures survive replays.

.changeset/chat-headstart-trigger-config.md

@@ -1,15 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Add `triggerConfig` support to `chat.headStart()` and `chat.openSession()`, so the auto-triggered handover-prepare run inherits tags, queue, machine, and other session trigger options the same way `chat.createStartSessionAction()` does. The `chat:{chatId}` tag is prepended automatically.
-
-```ts
-export const POST = chat.headStart({
-  agentId: "my-agent",
-  triggerConfig: { tags: ["org:acme"], queue: "chat" },
-  run: async ({ chat }) => streamText({ ...chat.toStreamTextOptions(), model }),
-});
-```
-
-Because the session is created once on the first head-start turn and is idempotent on the chat id, this is the only place to set those options for a head-start chat's lifetime. `chat.createStartSessionAction()` now also forwards `maxDuration`, `region`, and `lockToVersion` so both session entry points stay consistent.

.changeset/chat-history-read-primitives.md

@@ -1,21 +0,0 @@
----
-"@trigger.dev/sdk": minor
----
-
-Add read primitives to `chat.history` for HITL flows: `getPendingToolCalls()`, `getResolvedToolCalls()`, `extractNewToolResults(message)`, `getChain()`, and `findMessage(messageId)`. These lift the accumulator-walking logic that customers building human-in-the-loop tools were re-implementing into the SDK.
-
-Use `getPendingToolCalls()` to gate fresh user turns while a tool call is awaiting an answer. Use `extractNewToolResults(message)` to dedup tool results when persisting to your own store — the helper returns only the parts whose `toolCallId` is not already resolved on the chain.
-
-```ts
-const pending = chat.history.getPendingToolCalls();
-if (pending.length > 0) {
-  // an addToolOutput is expected before a new user message
-}
-
-onTurnComplete: async ({ responseMessage }) => {
-  const newResults = chat.history.extractNewToolResults(responseMessage);
-  for (const r of newResults) {
-    await db.toolResults.upsert({ id: r.toolCallId, output: r.output, errorText: r.errorText });
-  }
-};
-```

.changeset/chat-session-attributes.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"@trigger.dev/core": patch
----
-
-Stamp `gen_ai.conversation.id` (the chat id) on every span and metric emitted from inside a `chat.task` or `chat.agent` run. Lets you filter dashboard spans, runs, and metrics by the chat conversation that produced them — independent of the run boundary, so multi-run chats correlate cleanly. No code changes required on the user side.

.changeset/chat-slim-wire-merge.md

@@ -1,31 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Fix `chat.agent` HITL continuations on reasoning-heavy turns. Two changes that work together:
-
-- The per-turn merge now overlays the wire copy's tool-part state advancement onto the agent's existing chain — `state` + the matching resolution field (`output` / `errorText` / `approval`) come from the wire, everything else (text, reasoning, tool `input`, provider metadata) stays whatever the snapshot or `hydrateMessages` returned. Previously a full-message replace overwrote those fields with whatever the client shipped, so a slimmed wire copy landed a tool call with no `arguments` on the next LLM call. Covers `output-available` / `output-error` (HITL `addToolOutput`) and `approval-responded` / `output-denied` (approval flow).
-- `TriggerChatTransport.sendMessages` and `AgentChat.sendRaw` now slim assistant messages that carry advanced tool parts. The wire payload is just `{ id, role, parts: [<state + resolution field>] }` for `submit-message` continuations; everything else passes through. Reasoning blobs and full tool inputs no longer ride the wire on every `addToolOutput` / `addToolApproveResponse`, so continuation payloads stay well under the `.in/append` cap on long agent loops.
-
-Note: `onValidateMessages` receives the slim wire on HITL turns. If you call `validateUIMessages` from `ai` against the full `messages` array it will reject the slim assistant; filter to user messages (or skip on HITL turns) — see the updated docstring on `onValidateMessages` for the recommended pattern.
-
-For `hydrateMessages` hooks that persist the chain, this release also adds a small helper to the `@trigger.dev/sdk/ai` surface:
-
-```ts
-import { chat, upsertIncomingMessage } from "@trigger.dev/sdk/ai";
-
-chat.agent({
-  hydrateMessages: async ({ chatId, trigger, incomingMessages }) => {
-    const record = await db.chat.findUnique({ where: { id: chatId } });
-    const stored = record?.messages ?? [];
-    if (upsertIncomingMessage(stored, { trigger, incomingMessages })) {
-      await db.chat.update({ where: { id: chatId }, data: { messages: stored } });
-    }
-    return stored;
-  },
-});
-```
-
-It pushes fresh user messages by id, no-ops on HITL continuations (the incoming shares an id with the existing assistant — the runtime overlays the new tool-state advance), and skips on non-`submit-message` triggers. Returns `true` if it mutated `stored` so the caller knows whether to persist.
-
-Net effect: `chat.addToolOutput(...)` / `chat.addToolApproveResponse(...)` on multi-step reasoning agents (OpenAI Responses with `store: false`, Anthropic extended thinking, etc.) no longer blows the cap and no longer corrupts the LLM input.

.changeset/chat-start-session-action-typed-client-data.md

@@ -1,22 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Type `chat.createStartSessionAction` against your chat agent so `clientData` is typed end-to-end on the first turn:
-
-```ts
-import { chat } from "@trigger.dev/sdk/ai";
-import type { myChat } from "@/trigger/chat";
-
-export const startChatSession = chat.createStartSessionAction<typeof myChat>("my-chat");
-
-// In the browser, threaded from the transport's typed startSession callback:
-const transport = useTriggerChatTransport<typeof myChat>({
-  task: "my-chat",
-  startSession: ({ chatId, clientData }) =>
-    startChatSession({ chatId, clientData }),
-  // ...
-});
-```
-
-`ChatStartSessionParams` gains a typed `clientData` field — folded into the first run's `payload.metadata` so `onPreload` / `onChatStart` see the same shape per-turn `metadata` carries via the transport. The opaque session-level `metadata` field is unchanged.

.changeset/chat-start-session-api-client.md

@@ -1,13 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-`chat.createStartSessionAction` now accepts an `apiClient` option, so you can scope a chat session start to a specific environment's API config (`baseURL` / `accessToken`) without setting a global `TRIGGER_SECRET_KEY`. Useful when one server starts chats across more than one environment.
-
-```ts
-const startSession = chat.createStartSessionAction("my-chat", {
-  apiClient: { baseURL, accessToken },
-});
-
-await startSession({ chatId, clientData });
-```

.changeset/chat-system-prompt-caching.md

@@ -1,25 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Cache your chat agent's system prompt with Anthropic prompt caching. `chat.toStreamTextOptions()` now emits the system prompt as a cacheable message when you opt in, so a large, stable system block is billed at cache-read rates on every turn instead of full price.
-
-```ts
-// at the streamText call site (Anthropic sugar)
-streamText({
-  ...chat.toStreamTextOptions({ cacheControl: { type: "ephemeral" } }),
-  messages,
-});
-
-// provider-agnostic equivalent
-chat.toStreamTextOptions({
-  systemProviderOptions: { anthropic: { cacheControl: { type: "ephemeral" } } },
-});
-
-// or where the prompt is defined
-chat.prompt.set(SYSTEM_PROMPT, {
-  providerOptions: { anthropic: { cacheControl: { type: "ephemeral" } } },
-});
-```
-
-Without an option, `system` stays a plain string. Pairs with a `prepareMessages` cache breakpoint to cache the conversation prefix across turns too.

.changeset/chat-transport-recreate-missing-session.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-`useTriggerChatTransport` now recovers when restored session state points at a session that no longer exists in the current environment

.changeset/cli-deploy-skip-rewrite-timestamp.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-Add `TRIGGER_BUILD_SKIP_REWRITE_TIMESTAMP=1` escape hatch for local self-hosted builds whose buildx driver doesn't support `rewrite-timestamp` alongside push (e.g. orbstack's default `docker` driver).

.changeset/cli-dev-without-project.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-Running a CLI command like `dev`, `deploy`, `preview`, or `update` before initializing a project no longer crashes with a raw `Cannot find matching package.json` stack trace. The CLI now detects the missing project and points you to `npx trigger.dev@latest init` instead.

.changeset/cli-init-ai-tooling.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-`trigger init` now sets up your AI coding assistant as part of project setup: pick the MCP server, the agent skills, or both, then scaffold with the CLI or hand off to your assistant. Adds a new `getting-started` agent skill that teaches assistants how to bootstrap Trigger.dev (install the SDK, write `trigger.config.ts`, create a first task, run `trigger dev`), so the AI-driven setup path works end to end. It ships in the CLI alongside the existing skills, version-matched to your SDK.

.changeset/coerce-concurrency-key-to-string.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Coerce numeric `concurrencyKey` values to string at the API boundary across `tasks.trigger`, `tasks.batchTrigger`, and the Phase-2 streaming batch endpoint.

.changeset/create-session-stop-continuation.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Fix two `chat.createSession()` bugs: stopping a generation no longer wedges the run (the turn loop raced a `totalUsage` promise that never settles after a stop-abort), and continuation runs now wait for the next message instead of invoking the model with an empty prompt.

.changeset/custom-agent-loop-fixes.md

@@ -1,9 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Three fixes for custom agent loops (`chat.customAgent`, `chat.createSession`, and hand-rolled `MessageAccumulator` loops):
-
-- Continuation runs no longer replay already-answered user messages into the first turn. The `.in` resume cursor is now seeded before any listener attaches (the same boot logic `chat.agent` uses), so a chat that continues after a cancel, crash, or upgrade only sees genuinely new messages.
-- Steering a hand-rolled loop mid-stream no longer wipes the in-flight assistant response. `chat.pipeAndCapture` now stamps a server-generated message id on the stream, so a `prepareStep` injection keeps the partial text instead of replacing the message.
-- Task-backed tools (`ai.toolExecute`) now work from custom agent loops: the parent's session is threaded to the child run, so child tasks can stream progress into the chat with `chat.stream.writer({ target: "root" })` instead of failing with "session handle is not initialized".

.changeset/dequeue-latency-histogram.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Record client-side dequeue API latency in the supervisor consumer pool as a Prometheus histogram (`queue_consumer_pool_dequeue_duration_seconds`, labelled by `outcome`: success/empty/error).

.changeset/dev-branches.md

@@ -1,6 +0,0 @@
----
-"trigger.dev": patch
-"@trigger.dev/core": patch
----
-
-Add support for dev branches to the webapp and CLI. This allows humans (and agents) to run multiple local dev servers simultaneously, with a separate dashboard for each one.

.changeset/duplicate-task-ids.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/core": patch
-"trigger.dev": patch
----
-
-`dev` and `deploy` now fail with a clear error when two tasks are defined with the same id, including across different task types (e.g. a scheduled task and a regular task sharing an id). Previously the second definition silently overwrote the first, so one of the tasks would vanish with no warning. Task ids are detected as duplicates during indexing (naming each offending id and the files it was found in), and the same rule is enforced server-side when the background worker is registered.

.changeset/env-vars-tracing-forceflush-typecheck.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Fix `@trigger.dev/core` build: cast the underlying log record exporter when calling `forceFlush` so it typechecks against the updated OpenTelemetry `LogRecordExporter` type (which no longer declares `forceFlush`).

.changeset/envvars-import-is-secret.md

@@ -1,12 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-`envvars.upload` now accepts an optional `isSecret` flag, letting you create the imported variables as secret (redacted) environment variables. When omitted, variables default to non-secret.
-
-```ts
-await envvars.upload("proj_1234", "prod", {
-  variables: { STRIPE_SECRET_KEY: "sk_live_..." },
-  isSecret: true,
-});
-```

.changeset/errors-api-schemas.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Add request and response schemas for the new Errors API (error groups). These back the env-scoped HTTP endpoints for listing error groups, retrieving a single group, and changing its state (resolve, ignore, unresolve), plus a `filter[error]` option on the runs list to fetch the runs behind a group. Exported from `@trigger.dev/core/v3` so the SDK can reuse them.

.changeset/httpserver-skip-body-parsing.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Add an optional `skipBodyParsing` flag to the internal HTTP server route definition, letting a route respond without reading or parsing the request body.

.changeset/idempotency-key-catalog-eviction.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/core": patch
-"trigger.dev": patch
----
-
-Fix idempotency key metadata (original key + scope) being silently dropped when a single run creates more than 1000 idempotency keys. The in-process catalog that maps a key's hash back to its original key/scope is no longer bounded to 1000 entries, so `idempotencyKeys.create()` results retain their metadata regardless of how many are created in a run. The catalog is now cleared at each run boundary so it does not accumulate across warm-start runs.

.changeset/large-trigger-payload-offload.md

@@ -1,8 +0,0 @@
----
-"@trigger.dev/core": patch
-"@trigger.dev/sdk": patch
----
-
-Offload large trigger payloads to object storage before sending the trigger API request. The SDK uploads packets at or above the existing 128KB limit and sends an `application/store` pointer instead of embedding large JSON in the request body. `TriggerTaskRequestBody` now validates that `application/store` payloads are non-empty storage paths.
-
-Payload uploads use the same resolved `ApiClient` as the trigger call (including `requestOptions.clientConfig`), not only the global `apiClientManager.client` — so custom `baseURL`, access token, and preview branch apply to both presign and trigger.

.changeset/locals-key-dual-package-fix.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Fix `LocalsKey<T>` type incompatibility across dual-package builds. The phantom value-type brand no longer uses a module-level `unique symbol`, so a single TypeScript compilation that resolves the type from both the ESM and CJS outputs (which can happen under certain pnpm hoisting layouts) no longer sees two structurally-incompatible variants of the same type.

.changeset/mcp-agent-chat-sessions.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-The CLI MCP server's agent-chat tools (`start_agent_chat`, `send_agent_message`, `close_agent_chat`) now run on the new Sessions primitive, so AI assistants driving a `chat.agent` get the same idempotent-by-`chatId`, durable-across-runs behavior the browser transport gets. Required PAT scopes go from `write:inputStreams` to `read:sessions` + `write:sessions`.

.changeset/mcp-list-runs-region.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-MCP `list_runs` tool: add a `region` filter input and surface each run's executing region in the formatted summary.

.changeset/mcp-trigger-task-no-default-wait.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-The MCP server no longer tells the AI agent to wait for a run to complete after every `trigger_task` call. Waiting is now opt-in: the agent only waits when you ask it to (for example "trigger and then wait for it to finish"). This avoids burning tokens polling runs you didn't need to block on and keeps responses clearer.

.changeset/mint-token-command.md

@@ -1,9 +0,0 @@
----
-"trigger.dev": patch
----
-
-Adds `trigger.dev mint-token`, which mints a short-lived delegated token from your stored personal access token. The token authenticates against the API as you, can be narrowed with `--cap` and given a lifetime with `--ttl`, and prints to stdout so it can be captured.
-
-```bash
-UAT=$(trigger.dev mint-token --ttl 3600 --cap read:runs)
-```

.changeset/mock-chat-agent-test-harness.md

@@ -1,8 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"@trigger.dev/core": patch
----
-
-Unit-test `chat.agent` definitions offline with `mockChatAgent` from `@trigger.dev/sdk/ai/test`. Drives a real agent's turn loop in-process — no network, no task runtime — so you can send messages, actions, and stop signals via driver methods, inspect captured output chunks, and verify hooks fire. Pairs with `MockLanguageModelV3` from `ai/test` for model mocking. `setupLocals` lets you pre-seed `locals` (DB clients, service stubs) before `run()` starts.
-
-The broader `runInMockTaskContext` harness it's built on lives at `@trigger.dev/core/v3/test` — useful for unit-testing any task code, not just chat.

.changeset/mollifier-buffer-pipeline-list-entries.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/redis-worker": patch
----
-
-Pipeline the per-entry `HGETALL` fetches in `MollifierBuffer.listEntriesForEnv`. The previous serial implementation issued one Redis round-trip per runId returned by `LRANGE`, which dominated stale-sweep wall-time at any meaningful backlog (at the sweep's default maxCount=1000, this is ~1000 RTTs per env per pass). Behaviour is unchanged — entries are still skipped when the entry hash has been torn down by a concurrent drainer ack/fail between the LRANGE and the HGETALL.

.changeset/mollifier-configurable-constants.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/redis-worker": patch
----
-
-Make mollifier buffer and drainer internals configurable. `MollifierBuffer` now accepts `ackGraceTtlSeconds`, `maxRetriesPerRequest`, `reconnectStepMs`, and `reconnectMaxMs` options, and `MollifierDrainer` accepts `maxBackoffMs` and `backoffFloorMs`. All default to their previous hardcoded values, so existing behaviour is unchanged.

.changeset/mollifier-drain-batch-size.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/redis-worker": patch
----
-
-`MollifierDrainer` accepts a `drainBatchSize` option (default 1) that controls how many entries are popped per env per tick — in-flight handlers remain capped by the global `concurrency`. `MollifierBuffer` also gains `getDrainingCount()` / `listStaleDraining()`, backed by a new `mollifier:draining` ZSET maintained atomically with pop/ack/fail/requeue (observability-only).

.changeset/mollifier-redis-worker-primitives.md

@@ -1,9 +0,0 @@
----
-"@trigger.dev/redis-worker": patch
----
-
-Add MollifierBuffer and MollifierDrainer primitives for trigger burst smoothing.
-
-MollifierBuffer (`accept`, `pop`, `ack`, `requeue`, `fail`, `evaluateTrip`) is a per-env FIFO over Redis with atomic Lua transitions for status tracking. `evaluateTrip` is a sliding-window trip evaluator the webapp gate uses to detect per-env trigger bursts.
-
-MollifierDrainer pops entries through a polling loop with a user-supplied handler. The loop survives transient Redis errors via capped exponential backoff (up to 5s), and per-env pop failures don't poison the rest of the batch — one env's blip is logged and counted as failed for that tick. Rotation is two-level: orgs at the top, envs within each org. The buffer maintains `mollifier:orgs` and `mollifier:org-envs:${orgId}` atomically with per-env queues, so the drainer walks orgs → envs directly without an in-memory cache. The `maxOrgsPerTick` option (default 500) caps how many orgs are scheduled per tick; for each picked org, one env is popped (rotating round-robin within the org). An org with N envs gets the same per-tick scheduling slot as an org with 1 env, so tenant-level drainage throughput is determined by org count rather than env count.

.changeset/mollifier-tag-cap.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/redis-worker": patch
----
-
-Mollifier `mutateSnapshot` now enforces a tag cap: an `append_tags` patch carrying `maxTags` returns `"limit_exceeded"` (writing nothing) when the deduped tag count would exceed the limit, so a buffered run can't accumulate more tags via the tags API than the trigger validator allows at creation.

.changeset/otel-suite-0218.md

@@ -1,7 +0,0 @@
----
-"@trigger.dev/core": patch
-"trigger.dev": patch
-"@trigger.dev/sdk": patch
----
-
-Update the bundled OpenTelemetry packages to their latest releases (`@opentelemetry/sdk-node` 0.218.0, `@opentelemetry/core` 2.7.1, `@opentelemetry/host-metrics` 0.38.3).

.changeset/plugin-auth-path.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/plugins": patch
----
-
-The public interfaces for a plugin system. Initially consolidated authentication and authorization interfaces.

.changeset/pre.json

@@ -1,82 +0,0 @@
-{
-  "mode": "exit",
-  "tag": "rc",
-  "initialVersions": {
-    "coordinator": "0.0.1",
-    "docker-provider": "0.0.1",
-    "kubernetes-provider": "0.0.1",
-    "supervisor": "0.0.1",
-    "webapp": "1.0.0",
-    "@trigger.dev/build": "4.4.6",
-    "trigger.dev": "4.4.6",
-    "@trigger.dev/core": "4.4.6",
-    "@trigger.dev/plugins": "4.4.6",
-    "@trigger.dev/python": "4.4.6",
-    "@trigger.dev/react-hooks": "4.4.6",
-    "@trigger.dev/redis-worker": "4.4.6",
-    "@trigger.dev/rsc": "4.4.6",
-    "@trigger.dev/schema-to-json": "4.4.6",
-    "@trigger.dev/sdk": "4.4.6"
-  },
-  "changesets": [
-    "agent-skills-bundled-in-sdk",
-    "agent-skills",
-    "ai-prompts",
-    "ai-sdk-7-support",
-    "ai-tool-helpers",
-    "backpressure-scale-up-freeze",
-    "bundle-skills-single-pass",
-    "cap-idempotency-key-length",
-    "chat-agent-hardening",
-    "chat-agent-on-boot-hook",
-    "chat-agent-tools",
-    "chat-agent",
-    "chat-boot-cursor",
-    "chat-headstart-custom-backends",
-    "chat-headstart-hydrate",
-    "chat-headstart-reasoning",
-    "chat-headstart-trigger-config",
-    "chat-history-read-primitives",
-    "chat-session-attributes",
-    "chat-slim-wire-merge",
-    "chat-start-session-action-typed-client-data",
-    "chat-system-prompt-caching",
-    "chat-transport-recreate-missing-session",
-    "cli-deploy-skip-rewrite-timestamp",
-    "cli-dev-without-project",
-    "cli-init-ai-tooling",
-    "coerce-concurrency-key-to-string",
-    "create-session-stop-continuation",
-    "custom-agent-loop-fixes",
-    "dequeue-latency-histogram",
-    "duplicate-task-ids",
-    "env-vars-tracing-forceflush-typecheck",
-    "envvars-import-is-secret",
-    "large-trigger-payload-offload",
-    "locals-key-dual-package-fix",
-    "mcp-agent-chat-sessions",
-    "mcp-list-runs-region",
-    "mcp-trigger-task-no-default-wait",
-    "mock-chat-agent-test-harness",
-    "mollifier-buffer-pipeline-list-entries",
-    "mollifier-configurable-constants",
-    "mollifier-drain-batch-size",
-    "mollifier-redis-worker-primitives",
-    "mollifier-tag-cap",
-    "otel-suite-0218",
-    "plugin-auth-path",
-    "project-environments-endpoint",
-    "resource-catalog-runtime-registration",
-    "retry-middleware-errors",
-    "retry-sigsegv",
-    "runs-list-region-filter",
-    "s2-batch-transform-linger-fix",
-    "sessions-primitive",
-    "span-api-cached-cost",
-    "trigger-client",
-    "trigger-skill-namespace-and-docs",
-    "trigger-skills-installer",
-    "unflatten-attributes-conflict",
-    "warm-start-external-trace-context-leak"
-  ]
-}

.changeset/project-environments-endpoint.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Add `GetProjectEnvironmentsResponseBody` and `ProjectEnvironment` schemas for the new `GET /api/v1/projects/{projectRef}/environments` endpoint, which lists the parent environments (dev, staging, preview, prod) a personal access token can access for a project. Dev is scoped to the token owner and branch (preview child) environments are excluded.

.changeset/redis-worker-oldest-message-age.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/redis-worker": patch
----
-
-Add a `redis_worker.queue.oldest_message_age` observable gauge (unit `ms`, labeled `worker_name`) reporting the age of the oldest overdue message in each queue. This is a generic queue-stall signal: it stays at 0 while a queue drains healthily and rises only when due work sits undrained (e.g. a blocked dequeue, a dead consumer, or backpressure), even when no items are being processed. Orphaned queue entries are resolved against the items hash so they don't report a phantom stall. Also exposes `SimpleQueue.oldestMessageAge()`.

.changeset/resource-catalog-runtime-registration.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/core": patch
-"trigger.dev": patch
----
-
-Fix `COULD_NOT_FIND_EXECUTOR` when a task's definition is loaded via `await import(...)` from inside another task's `run()`. The runtime workers now register such tasks with a sentinel file context, and the catalog logs a one-time warning per task id.

.changeset/retry-middleware-errors.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Retry `TASK_MIDDLEWARE_ERROR` under the task's retry policy instead of failing the run on the first attempt. The error was already classified as retryable by `shouldRetryError`, but `shouldLookupRetrySettings` did not include it, so the retry flow fell through to `fail_run`. Fixes #3231.

.changeset/retry-sigsegv.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Retry `TASK_PROCESS_SIGSEGV` task crashes under the user's retry policy instead of failing the run on the first segfault. SIGSEGV in Node tasks is frequently non-deterministic (native addon races, JIT/GC interaction, near-OOM in native code, host issues), so retrying on a fresh process often succeeds. The retry is gated by the task's existing `retry` config + `maxAttempts` — same path `TASK_PROCESS_SIGTERM` and uncaught exceptions already use — so tasks without a retry policy still fail fast.

.changeset/runner-send-debug-logs-gate.md

@@ -1,5 +0,0 @@
----
-"trigger.dev": patch
----
-
-Runner debug logs are now disabled by default. Set `SEND_RUN_DEBUG_LOGS=true` on the supervisor to re-enable them.

.changeset/runs-list-region-filter.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/core": patch
-"@trigger.dev/sdk": patch
----
-
-Add `region` to the runs list / retrieve API: filter runs by region (`runs.list({ region: "..." })` / `filter[region]=<masterQueue>`) and read each run's executing region from the new `region` field on the response.

.changeset/s2-batch-transform-linger-fix.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/core": patch
-"trigger.dev": patch
----
-
-Bump `@s2-dev/streamstore` to `0.22.10` to fix a `TASK_RUN_UNCAUGHT_EXCEPTION` ("Invalid state: Unable to enqueue") when a `chat.agent` turn is aborted mid-stream.

.changeset/sessions-primitive.md

@@ -1,26 +0,0 @@
----
-"@trigger.dev/sdk": minor
-"@trigger.dev/core": patch
----
-
-**Sessions** — a durable, run-aware stream channel keyed on a stable `externalId`. A Session is the unit of state that owns a multi-run conversation: messages flow through `.in`, responses through `.out`, both survive run boundaries. Sessions back the new `chat.agent` runtime, and you can build on them directly for any pattern that needs durable bi-directional streaming across runs.
-
-```ts
-import { sessions, tasks } from "@trigger.dev/sdk";
-
-// Trigger a task and subscribe to its session output in one call
-const { runId, stream } = await tasks.triggerAndSubscribe("my-task", payload, {
-  externalId: "user-456",
-});
-
-for await (const chunk of stream) {
-  // ...
-}
-
-// Enumerate existing sessions (powers inbox-style UIs without a separate index)
-for await (const s of sessions.list({ type: "chat.agent", tag: "user:user-456" })) {
-  console.log(s.id, s.externalId, s.createdAt, s.closedAt);
-}
-```
-
-See [/docs/ai-chat/overview](https://trigger.dev/docs/ai-chat/overview) for the full surface — Sessions powers the durable, resumable chat runtime described there.

.changeset/span-api-cached-cost.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-The run span API response now includes `cachedCost` and `cacheCreationCost` on the `ai` object, alongside the existing `inputCost` / `outputCost` / `totalCost`. `inputCost` reflects only the non-cached input, so these fields let you reconstruct the full cost breakdown for prompt-cached calls.

.changeset/tidy-exec-arg-logging.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Redact credential-bearing flag values (e.g. `--password`, `--token`) from `Exec` command debug logs

.changeset/trigger-client.md

@@ -1,18 +0,0 @@
----
-"@trigger.dev/sdk": patch
----
-
-Add `TriggerClient` for running multiple SDK clients side-by-side, each with its own auth, preview branch, and baseURL. Useful when a single process needs to trigger tasks or read runs across multiple projects, environments, or preview branches without mutating shared global state.
-
-```ts
-import { TriggerClient } from "@trigger.dev/sdk";
-
-const prod = new TriggerClient({ accessToken: process.env.TRIGGER_PROD_KEY });
-const preview = new TriggerClient({
-  accessToken: process.env.TRIGGER_PREVIEW_KEY,
-  previewBranch: "signup-flow",
-});
-
-await prod.tasks.trigger("send-email", payload);
-await preview.runs.list({ status: ["COMPLETED"] });
-```

.changeset/trigger-skill-namespace-and-docs.md

@@ -1,6 +0,0 @@
----
-"@trigger.dev/sdk": patch
-"trigger.dev": patch
----
-
-The agent skills installed by `trigger skills` are now namespaced with a `trigger-` prefix (e.g. `trigger-authoring-tasks`, `trigger-getting-started`) so they don't collide with unrelated skills in your coding agent's skills directory. Adds a `trigger-cost-savings` skill for auditing and reducing compute spend (right-sizing machines, `maxDuration`, batching, debounce), and `@trigger.dev/sdk` now bundles the full Trigger.dev documentation so your agent can read the complete, version-pinned reference directly from node_modules.

.changeset/trigger-skills-installer.md

@@ -1,11 +0,0 @@
----
-"trigger.dev": patch
----
-
-`trigger skills` installs Trigger.dev agent skills into your coding agent so it knows how to write tasks, schedules, realtime, and chat.agent code. The skills ship with the CLI and are copied into each tool's native skills directory (Claude Code, Cursor, GitHub Copilot, and Codex / AGENTS.md), and `trigger dev` offers to install them on first run.
-
-```bash
-trigger skills --target claude-code
-```
-
-Replaces the previous `install-rules` command, which stays as an alias.

.changeset/unflatten-attributes-conflict.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Fix `TypeError` in `unflattenAttributes` when the input attribute map contains conflicting dotted key paths (e.g. both `a.b` set to a scalar and `a.b.c` set to a value). The path-walk loop now applies last-write-wins when a prior key wrote a primitive, null, or array at an intermediate slot, matching the existing precedent in `AttributeFlattener.addAttribute`. Callers no longer crash when handed malformed external attribute inputs.

.changeset/warm-start-external-trace-context-leak.md

@@ -1,5 +0,0 @@
----
-"@trigger.dev/core": patch
----
-
-Fix external trace context leaking across runs on warm-started workers with `processKeepAlive` enabled. Every subsequent run's attempt span was being exported with the first run's `traceId` and `parentSpanId`, breaking causal-chain navigation in external APM tools. Runs without an external trace context are unaffected.

.server-changes/billing-limits.md

@@ -1,24 +0,0 @@
----
-area: webapp
-type: feature
----
-
-Add billing limits. Customers set a spend cap; when usage crosses it, billable
-environments pause for a grace period, new triggers are rejected once it ends,
-and a recovery flow resumes or cancels the queued backlog. Reconciliation keeps
-the webapp converged to billing's state.
-
-## Manual pause during billing enforcement
-
-While `pauseSource=BILLING_LIMIT`, manual resume is rejected and manual pause is
-a silent no-op (`PauseEnvironmentService` returns success with state `paused`).
-We do not stack a manual pause on top of billing enforcement because resolve
-converge unpauses all `BILLING_LIMIT`-paused environments for the org.
-
-API callers that pause during enforcement should expect the environment to
-resume when the billing limit is resolved. The queues UI hides pause/resume in
-this state; see `manualPauseEnvironmentGuard.server.ts`.
-
-The admin `runs.enable` endpoint skips billing-paused environments when
-re-enabling or disabling org runs (returns them in `skipped`, not `failures` or
-the update count). They resume only after the billing limit is resolved.

.server-changes/bulk-replay-region-override.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: feature
----
-
-Add an "Override region" option to the bulk replay action so replayed runs can be routed to a chosen region, defaulting to keeping each run in its original region.

.server-changes/cached-task-icon-svg.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Refresh the task and cached-task span icons shown in the run trace view with new SVG artwork.

.server-changes/clickhouse-array-of-dynamic-inference.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Infer mixed-type JSON arrays as Array(Dynamic) instead of nested tuples when writing run, event, metric, and session data to avoid ClickHouse type-complexity merge failures

.server-changes/conform-v1.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Upgrade the dashboard form layer from `@conform-to` 0.9 to 1.x. conform 1.x supports both zod 3 and zod 4, which unblocks the upcoming zod 4 upgrade.

.server-changes/custom-chart-categorical-x-axis.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Custom (Query mode) and dashboard charts keep non-date x-axis labels like run IDs and task names readable with no configuration: width-aware label thinning, middle-truncation with the full value on hover, and auto-rotation only when labels are long.

.server-changes/dashboard-line-chart-x-axis-density.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Dashboard and custom query line charts now choose how many x-axis time labels to show based on the chart's rendered width, so wide charts show more labels and narrow widgets show fewer.

.server-changes/dequeue-region-gate.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: feature
----
-
-Add a `RUN_ENGINE_DEQUEUE_DISABLED_WORKER_QUEUES` setting that refuses worker dequeue requests for the listed worker queues (or base regions), so their runs stay queued instead of being handed to workers that can't run them.

.server-changes/dev-branches.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: feature
----
-
-Adds support for dev branches similar to the preview branches already supported.

.server-changes/fix-invite-accept-many-projects.md

@@ -1,8 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Fixed invite acceptance failing for organizations with many projects.
-
-When environment provisioning failed after membership was created, users with a single pending invite were redirected away before seeing the error. They now land on the orgs page with a persistent error toast; users with other pending invites still see a FormError on the invites page.

.server-changes/llm-spend-currency-label.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Add a currency unit to the agent dashboard "LLM spend" chart label, so it now reads "LLM spend ($)".

.server-changes/logs-search-memory-and-pagination.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Keep logs search within bounded ClickHouse memory when browsing long time ranges, and fix pagination that could skip or duplicate entries sharing a timestamp.

.server-changes/prisma-infrastructure-error-capture.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Log Prisma infrastructure errors (P1xxx) centrally and obfuscate their messages (which carry the DB hostname) on API responses that previously returned the raw message, without changing status codes or headers.

.server-changes/rbac-permission-enforcement.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: feature
----
-
-Enforce role-based permissions across the dashboard and API. New permission boundaries cover: runs (cancel, replay, bulk actions), deployments (rollback, promote, cancel), prompt versions, organization members (invite, resend, revoke), billing and seat purchases, integrations (GitHub and Vercel), and environment variables and API keys (restricted by environment tier). Roles without access can no longer read or change these, gated controls are disabled with a tooltip, and gated pages show a permission-denied panel instead of redirecting away. Behaviour is unchanged in the default configuration, where permissions stay permissive.

.server-changes/remove-worker-create-endpoint.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: breaking
----
-
-Remove the unused worker group management API endpoints (GET and POST /api/v1/workers).

.server-changes/route-taskrun-reads-through-run-store.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Route Postgres task run reads through the run store so they can be retargeted to a different backing store without changing call sites.

.server-changes/scheduled-task-out-of-entitlements-warning.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Treat a scheduled task trigger that fails because the organization is out of entitlements as an expected outcome: the schedule engine now logs it as a warning instead of an error, mirroring how environment queue-limit results are already handled.

.server-changes/sessions-test-column.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: feature
----
-
-Agent sessions started from the Test playground are now flagged with a real `Session.isTest` boolean instead of a `"playground"` tag, surfaced as a dedicated "Test" column (check icon) in the Sessions table on both the Sessions and Agent pages, plus a matching property on the session detail page. The legacy `"playground"` tag is hidden from the Tags display on pre-existing sessions.

.server-changes/sso.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: feature
----
-
-SAML/OIDC single sign-on: SSO login with optional per-domain enforcement, JIT provisioning, and periodic re-validation against the IdP.

.server-changes/supervisor-pod-count-backpressure.md

@@ -1,6 +0,0 @@
----
-area: supervisor
-type: feature
----
-
-The supervisor can pause dequeuing when the Kubernetes cluster is saturated, based on the cluster's total pod count. Opt-in and off by default.

.server-changes/swap-task-icons.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Update the dashboard task icons with a new glyph design.

.server-changes/task-landing-activity-charts.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Improve the activity charts on the task landing pages (agent, standard, scheduled): bar density now adapts to the selected time range so short ranges no longer collapse to a single bar, x-axis labels are width-aware and non-overlapping, the agent charts share a synced hover line, each chart gets a maximize button, and dragging across a chart zooms the Time/Date filter.

.server-changes/task-run-plan-type-clickhouse.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Store the run's plan type on the runs analytics table so reporting can group runs by plan.

.server-changes/task-type-filter-segmented-control.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-Replace the Task type filter on the Tasks page with a segmented control: "All" plus icon-only Agent, Standard, and Scheduled segments (each with a tooltip showing its label and number-key shortcut). Filtering is now single-select (one task type at a time) instead of multi-select. Shortcut keys 0–3 select each segment.

.server-changes/tasks-page-hydration-errors.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Stop the Tasks page from logging React hydration errors for the per-row running and activity stats.

.server-changes/toast-message-length-clamp.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Dashboard error toasts with very long messages no longer exceed the session cookie limit and break the request; over-long messages are truncated.

.server-changes/trace-view-large-runs-subtree.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: fix
----
-
-Fix empty trace views for child and nested runs in very large traces. The dashboard and retrieve-trace API now return the requested run's span subtree, including ancestor spans outside the anchor run's time window (so a parent's cancellation/error state propagates down correctly).

.server-changes/v3-engine-retirement-messaging.md

@@ -1,6 +0,0 @@
----
-area: webapp
-type: improvement
----
-
-When the v3 engine is retired, triggering a v3 task and connecting the v3 dev CLI now fail with a clear message pointing to the v4 migration guide instead of failing opaquely. Enforcement is off by default, so self-hosted instances still running v3 are unaffected until they migrate.