Feat/agent mode#28
Merged
Merged
Conversation
- Add Peggy parser build step to package.json scripts - Add tsconfig path aliases for @thinkix/ai sub-exports - Update vitest aliases to match new AI package layout - Configure Playwright to run against production build on port 3100 - Exclude generated components/ dir and parser.js from linting - Upgrade ai SDK to v6 and add agent-related dependencies
- Move lib/ -> src/ with updated exports - Replace canvas command model with Zod tool schemas - Add provider resolution with env var fallback chain - Add system prompt builder and DSL reference - Add web search tool definition
- Add STICKY_COLORS map and DEFAULT_STICKY_COLOR constant - Add createStickyNoteElement helper with color support - Refactor with-sticky-note plugin to use shared color config
- Add calculateFocusedViewport for zoom-to-fit on new elements - Add focusAndRevealElements with animated pan/zoom and reveal effects - Add insertElementDirect for single-element board insertion - Export new utilities from barrel
- Add AgentPane component with Cmd+J toggle and resizable panel - Add AgentToolRenderer for custom tool result display - Add AgentSettingsDialog for per-user API key configuration - Add useAgentChat hook with streaming and tool execution - Add Peggy-based DSL parser (thinkix.peggy) for canvas operations - Add DSL compiler converting AST to Plait elements - Add serializer converting Plait elements back to DSL - Add file-system-ops for ls, read, write, patch, rm, select tools
- Add /api/agent route with streaming LLM responses via AI SDK v6 - Add /api/agent/config endpoint for safe client-side config - Remove deprecated /api/chat and /api/structure routes
- Extract boardAdapter and BoardInfo/BoardData types - Add mapBoardDtoToBoard and toBoardMetadata helpers - Return next board after delete instead of void - Sync board metadata on save operations - Add use-board-store unit tests
- Add initialElements and roomId props to MockYjsProvider - Derive syncState from browser online/offline events - Export YjsCollaborationContext and context value type - Extract localStorage helpers in use-collaboration hook - Immediately persist enable/disable actions to storage - Remove unused MockRoom component
- Bridge board state between Plait and Yjs mock provider - Pass roomId and initialElements to MockYjsProvider - Sync local changes to collaboration and vice versa - Properly disable collaboration and clear stored state
- Add Cmd+J keyboard shortcut to toggle agent pane - Render AgentPane in resizable side panel - Add data-element-count and data-has-elements attrs to board div - Minor whitespace cleanup in BoardLayoutSlots
- Add testids to AppMenu items, export options, and dialogs - Add testid to shape tool menu items - Add testids to collaborate button and status bar elements - Add dataTestId prop to ShareButton component
- Register global keydown listener with editor/input exclusion - Support both Cmd+Z and Cmd+Shift+Z / Cmd+Y for redo - Move board null check after hook to satisfy rules-of-hooks - Add data-testid to undo and redo buttons
- Add streamdown styles import for markdown rendering - Add CSS custom property mappings for theme colors - Fix plait text container wrapping with pre-wrap and word-break - Add lib/utils.ts with cn() utility
- Handle getActors returning object instead of array in sequence parser - Replace @thinkix/shared logger with inline console logger in flowchart - Preserve metadata from text nodes in class transformer
- Switch MarkdownToMindmapDialog to insertElementDirect + focus/reveal - Add focus/reveal animation after mermaid board insertion - Separate preview parsing from insert parsing in markdown dialog
- Centralize board wait, app menu, and collab helpers in utils.ts - Replace fragile text/role selectors with data-testid - Use production build (port 3100) instead of dev server - Simplify collaboration tests with enable/draw helpers - Add agent pane E2E spec and agent stream test helper
- Remove old ai-commands test (replaced by tool schemas) - Expand ai-core tests for provider resolution and env fallbacks - Add tests for DSL compiler, serializer, system prompt, tool schemas - Add tests for run-tool, connection-graph, diagram-layout, presentation-layer - Add component tests for agent-tool-renderer and diagram-preview - Update insertion and sticky-note tests for new utilities - Update mermaid and MermaidToBoardDialog tests
- Add AGENTS.md with comprehensive project guide for Codex - Update CLAUDE.md with agent architecture and DSL reference - Update README with server env var configuration - Add .env.example with all supported AI env vars
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Contributor
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 18 minutes and 31 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (2)
📒 Files selected for processing (131)
📝 WalkthroughWalkthroughThis pull request introduces a comprehensive AI agent system for Thinkix that enables AI-driven board inspection and manipulation through a custom DSL. It adds a streaming agent API, client-side chat UI with tool rendering, a domain-specific language parser/compiler for board commands, extensive board operation utilities, and substantial new documentation and test coverage. Changes
Sequence Diagram(s)sequenceDiagram
participant Client as Client (Browser)
participant APIDest as /api/agent
participant AIModel as AI Model
participant BoardState as Board State
participant Tool as Tool Executor
participant Storage as Storage/Board
Client->>APIDest: POST {messages, provider, model, apiKey, context}
APIDest->>APIDest: Resolve provider/model/key config
APIDest->>APIDest: Convert UI messages to model format
APIDest->>AIModel: streamText() with system prompt & tools
loop Streaming Response
AIModel-->>APIDest: Text chunks
APIDest-->>Client: Text UI message events
alt Tool Call Detected
AIModel-->>APIDest: {toolName: 'run', input: {command}}
APIDest->>Tool: executeRun(command, boardOptions)
Tool->>BoardState: Parse DSL & execute commands
Tool->>Tool: Apply board mutations (touch/write/patch/etc)
Tool->>Storage: Persist updated board
Tool-->>APIDest: {kind, summary, data}
APIDest-->>Client: Tool-input + tool-result events
end
end
Client->>Client: Render streamed messages + tool outputs
sequenceDiagram
participant DSL as DSL Input
participant Parser as PEG Parser
participant Compiler as Compiler
participant Result as Board Result
DSL->>Parser: "shape rect 'Box' id:myshape"
Parser->>Parser: Apply grammar rules
Parser-->>Compiler: {statements: [ShapeNode]}
Compiler->>Compiler: Dispatch by node.type
Compiler->>Compiler: Create geometry via SHAPE_MAP
Compiler->>Compiler: Apply text, color, position
Compiler-->>Result: {elements: [PlaitElement], ...}
Result->>Result: Insert into board + apply layout
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes The diff introduces substantial new logic across multiple domains: DSL implementation with a Peggy grammar and generated parser, a complex command execution engine with 13+ board operations, server-side streaming architecture with tool calling, and extensive client-side UI components. While many changes follow consistent patterns (e.g., repeated test files, similar UI component structures), the core DSL/command/streaming systems require careful validation of control flow, state mutations, and integration points. The heterogeneous nature—spanning API routes, UI components, data structures, build configuration, and tests—increases cognitive load despite individual pieces being moderate in complexity. Possibly related PRs
Poem
✨ Finishing Touches🧪 Generate unit tests (beta)
|
❌ Test Results
How to view coverage report
Commit: |
This was referenced Apr 6, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary by CodeRabbit
Release Notes
New Features
Bug Fixes & Improvements
Documentation