refactor: v2 - ai assistant - introduce agentthread primitive

[maxy-shpfy]

Description

The AI chat agent client has been refactored from a global singleton into a per-thread model. A new AgentThread primitive encapsulates a single conversation's Web Worker, MobX state (messages, thinking text, pending status), and lifecycle. AiChatStore now owns a collection of AgentThread instances and tracks an active thread ID, with newThread() disposing the previous thread and spinning up a fresh one.

A "New chat" button has been added to the chat UI header, allowing users to explicitly start a fresh conversation. The AiChatStoreProvider now calls ensureActiveThread() on mount and disposeAll() on unmount, correctly handling React StrictMode's mount → unmount → mount cycle which previously left the store in an empty state.

The threadId is now owned by AgentThread and injected into every worker request directly, rather than being tracked as a response value and passed back on subsequent messages.

Related Issue and Pull requests

Type of Change

• Bug fix
• New feature
• Improvement
• Cleanup/Refactor
• Breaking change
• Documentation update

Checklist

• I have tested this does not break current pipelines / runs functionality
• I have tested the changes on staging

Screenshots (if applicable)

Demo of AI multi thread.mov (uploaded via Graphite)

Test Instructions

1. Open the AI chat panel and verify a conversation can be sent and receives a response as normal.
2. Click the "New chat" (pencil) button in the chat header and confirm the message history is cleared and a fresh conversation begins.
3. Verify that navigating away and back to the chat panel does not leave the store in an empty/broken state.
4. Confirm that under React StrictMode (development builds) the chat initialises correctly after the double-mount cycle.

Additional Comments

The AgentThread class includes a placeholder comment for a future persist() method to support snapshotting and restoring threads across page reloads, which is out of scope for this change.

[maxy-shpfy]

• feat: v2 - tangent - connect dashboard to experiment scenarios #2387
• feat: v2 - tangent - progress update poll #2386
• feat: v2 - tangent - persist url to research session in opencode agent #2371
• feat: v2 - tangent - facelift for run details panel #2369
• feat: v2 - tangent - suggest optimization button #2368
• feat: v2 - tangent - optimization scenario card condensed #2367
• fix: v2 - tangent - cors worker load #2364
• feat: v2 - tangent - autoresearch run #2363
• feat: v2 - tangent - scenarios window #2362
• feat: v2 - ai assistant - tangent researcher subagent #2361
• poc: scaffold basic tangent dashboard ux #2360
• chore: generate api sdk from tangent #2359
• refactor: v2 - run view - run menu #2358
• refactor: v2 - run view - windows top menu #2357
• refactor: v2 - ai assistant - split editor and view agent implementations #2356
• refactor: v2 - ai assistant - introduce agentthread primitive #2355 👈 (View in Graphite)
• refactor: v2 - ai assistant - basic chat in RunView #2354
• refactor: v2 - ai assistant - move ai chat components to shared space #2353
• fix: v2 - window state for startVisible #2352
• master

This stack of pull requests is managed by Graphite. Learn more about stacking.

[github-actions]

🎩 Preview

A preview build has been created at: 06-02-refactor_v2_-_ai_assistant_-_introduce_agentthread_primitive/a19dc61

[maxy-shpfy]

Merge activity

• Jun 10, 11:59 PM UTC: A user started a stack merge that includes this pull request via Graphite.
• Jun 11, 12:04 AM UTC: Graphite couldn't merge this pull request because a downstack PR refactor: v2 - ai assistant - move ai chat components to shared space #2353 failed to merge.
• Jun 11, 6:05 AM UTC: A user started a stack merge that includes this pull request via Graphite.
• Jun 11, 6:13 AM UTC: Graphite rebased this pull request as part of a merge.
• Jun 11, 6:16 AM UTC: @maxy-shpfy merged this pull request with Graphite.