Dashboard Agent V1 — chat, reports, Investigate, Watch

.changeset/report-json-and-period-units.md

@@ -0,0 +1,6 @@
+---
+"@trigger.dev/core": patch
+"trigger.dev": patch
+---
+
+Reports can now be fetched as structured data, not just text: ask for the `json` format and you get the numbers and what they mean, typed. Report periods are also stricter — the shortest window is one minute (`30m`, `1h`, `7d`), because reports summarise data by the minute and anything shorter can't be answered honestly.

.server-changes/ask-agent-replaces-ask-ai.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: improvement
+---
+
+Ask AI has been replaced by the dashboard agent everywhere it used to appear — the Help & Feedback menu, the deployment setup panels, and the link the CLI prints when a task errors. Chat answers now read as plain text instead of stacked cards, and on the Free plan you get 20 messages before an upgrade is needed (your past chats stay readable).

.server-changes/dashboard-agent-first-turn-error.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: fix
+---
+
+When the assistant fails to start its very first reply, the chat now shows an error you can retry instead of sitting empty.

.server-changes/dashboard-agent-hero-and-fullscreen.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: improvement
+---
+
+A new chat with the agent now opens with a bigger prompt field and one-click suggestions for the page you're on. You can also expand the agent to fill the screen and collapse it back to the side panel, keeping your conversation.

.server-changes/dashboard-agent-investigate.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+The dashboard agent can now investigate failures for you: click Investigate on a failed run, an error, a backed-up queue, or a run that hasn't started yet and it gathers the evidence, tests a few hypotheses, and answers with what happened and how to fix it — every claim linked to the runs, errors, and deploys behind it. Available to organizations with the dashboard agent enabled.

.server-changes/dashboard-agent-prompts-all-pages.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+The dashboard agent's suggested questions now fit every page in an environment — tasks, schedules, batches, alerts, waitpoints, bulk actions, branches, limits, dashboards, agents, prompts, models, sessions and the rest — instead of falling back to generic ones. Pages where something is actually stuck offer a question about it: a scheduled task with no schedule attached, a paused queue, a batch whose runs failed, a wait token past its timeout, a prompt pinned to an override.

.server-changes/dashboard-agent-prompts-more-pages.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+The dashboard agent's suggested questions now fit the page you're on across errors, queues and deployments, not just runs — an error group, a backed-up queue or a failed deploy each offer questions about that specific thing.

.server-changes/dashboard-agent-stuck-investigations.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: fix
+---
+
+An investigation card no longer spins forever when the agent doesn't come back to it. It now closes on its own as inconclusive, keeping everything that was checked.

.server-changes/dashboard-agent-watch-alerts.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+Watches you set up with the dashboard agent can now alert you by email, Slack, or webhook when they fire. Pick the new "Dashboard agent watches" type on the Alerts page, and turn it off again from any alert email.

.server-changes/dashboard-agent-watch-card.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+There's now a **Watch…** button on runs, queues, errors and the health report. It opens a short form with the right thing to wait for already filled in — a run finishing, a queue clearing, an error coming back, an environment recovering — so one click is enough. Open **Customize** first if you'd rather change how long it waits, how often it checks, or what it waits for, and you can ask for an email as well as the chat message.

.server-changes/dashboard-agent-watch-investigates-itself.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: improvement
+---
+
+When you ask a watch to investigate bad news, it now does the whole thing on its own: the chat tells you the watch resolved, and the findings arrive right after in their own message, with the investigation card. You no longer have to come back and ask.

.server-changes/dashboard-agent-watch-queue-conditions.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+A queue watch can now wait for three more things under **Customize**: the queue coming back below a number you pick, the queue stopping moving at all, and runs waiting longer than a limit you set. On a queue where runs are already waiting too long, the **Watch…** button opens on that wait instead of on "until it clears".

.server-changes/dashboard-agent-watches.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+Ask the dashboard agent to tell you when something happens — a run starting or finishing, a queue clearing or growing past a number you pick, an error coming back, an environment recovering — and it messages you in the chat once with the answer. It tells you either way: that the run finished, that it failed, or that the queue still hadn't cleared by the time it stopped looking. If the thing you asked about has already happened, it just says so instead of waiting. Each chat can wait on up to three things at a time, for up to 24 hours. You can also ask it to start looking into the cause if the news turns out to be bad, and it will — otherwise it just tells you and stops.

.server-changes/queue-metrics-api.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: feature
+---
+
+You can now read a single queue's wait times, peak depth, throughput, and how often it was throttled over a time window from the API.

.server-changes/remove-header-docs-buttons.md

@@ -0,0 +1,6 @@
+---
+area: webapp
+type: improvement
+---
+
+The Docs button has been removed from page headers to reduce clutter. Documentation links remain available in context where they're most useful.

apps/webapp/.gitignore

@@ -7,6 +7,9 @@ node_modules
 /cypress/screenshots
 /cypress/videos
 
+# Output of `pnpm run agent-ui:screenshots`
+/screenshots
+
 /app/styles/tailwind.css
 
 # Ensure the .env symlink is not removed by accident
@@ -20,4 +23,6 @@ storybook-static
 
 /prisma/seed.js
 /prisma/populate.js
-.memory-snapshots
+.memory-snapshots
+# Heartbeat story mode for seed-agent-examples (degraded | calm)
+.agent-examples-heartbeat-mode