Fixes#1405
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Adds GPT-5 Codex (OpenAI and Azure) and Claude 4.5 Sonnet to the model
options to enable newer coding models and larger contexts. Also
increases Claude 4 Sonnet max output tokens to 32k.
<!-- End of auto-generated description by cubic. -->
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Adds GPT‑5 Codex (OpenAI/Azure) and Claude 4.5 Sonnet, and increases
Claude 4 Sonnet max output tokens to 32k across providers and tests.
>
> - **Models**:
> - **OpenAI**: add `gpt-5-codex` (400k context, default temp 1).
> - **Anthropic**:
> - add `claude-sonnet-4-5-20250929` (1M context, `maxOutputTokens:
32_000`).
> - update `claude-sonnet-4-20250514` `maxOutputTokens` from `16_000` to
`32_000`.
> - **Azure**: add `gpt-5-codex` (400k context, `maxOutputTokens:
128_000`).
> - **Bedrock**:
> - add `us.anthropic.claude-sonnet-4-5-20250929-v1:0` (1M context,
`maxOutputTokens: 32_000`).
> - update `us.anthropic.claude-sonnet-4-20250514-v1:0`
`maxOutputTokens` to `32_000`.
> - **E2E tests**:
> - Update snapshots to reflect `max_tokens` increased to `32000` for
`anthropic/claude-sonnet-4-20250514` in engine and gateway tests.
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
73298d2da0c833468f957bb436f1e33400307483. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
Fixes#912
This PR implements disabling send button while approval is pending and
addresses issue #912
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Disable the chat send button while a proposal is awaiting approval, and
re-enable it after approve or reject. Prevents accidental messages
during pending changes. Addresses issue #912.
- **New Features**
- Track pending changes with isChangesPending based on the last
assistant message’s approvalState.
- Disable the send button when a proposal is pending (in addition to the
existing empty input check).
- Re-enable after approve/reject by refreshing the proposal and
messages.
- Added Playwright e2e tests for both approve and reject flows.
<!-- End of auto-generated description by cubic. -->
<!-- CURSOR_SUMMARY -->
---
> [!NOTE]
> Disable the chat send button when a code proposal is pending approval
and re-enable after approve/reject; add e2e coverage and update MCP
flow.
>
> - **Frontend**
> - `ChatInput.tsx`: Read `messages` from `chatMessagesAtom` and derive
`disableSendButton` when the last assistant message (matching
`proposal.messageId`) has no `approvalState` and `proposal.type ===
"code-proposal"`.
> - Apply `disableSendButton` to the send button’s `disabled` condition
(in addition to empty input/attachments).
> - Ensure proposal/messages refresh after approve/reject.
> - **Tests**
> - Add Playwright tests `e2e-tests/chat_input.spec.ts` to verify send
button disabled during pending proposal and re-enabled after approve or
reject.
> - Update `e2e-tests/mcp.spec.ts` to click "Approve" after granting
consent.
>
> <sup>Written by [Cursor
Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit
b9b47bd6f547449cc5cf1d39a00e4e7fb5de1bcd. This will update automatically
on new commits. Configure
[here](https://cursor.com/dashboard?tab=bugbot).</sup>
<!-- /CURSOR_SUMMARY -->
---------
Co-authored-by: Will Chen <willchen90@gmail.com>
## Summary
Adds AI response copy functionality to chat messages that preserves
formatting and converts Dyad-specific tags to clean, readable markdown.
## Changes
- **New `useCopyToClipboard` hook**: Parses Dyad tags and converts them
to professional markdown format
- **Updated `ChatMessage` component**: Positions copy button on left
side of approval status
- **Dyad tag conversion**: Transforms custom tags to readable format:
- `<dyad-write>` → `### File: path/to/file.js` + code block
- `<dyad-edit>` → `### Edit: path/to/file.js` + code block
- `<dyad-execute-sql>` → `### Execute SQL` + ```sql block
- `<think>` → `### Thinking` + content
## Features
- ✅ Automatic programming language detection from file extensions
- ✅ Professional markdown formatting with proper headings and code
blocks
- ✅ Tooltip showing "Copied" confirmation
- ✅ Reuses existing DyadMarkdownParser logic for consistency
closes (#1290)
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Adds a Copy button to assistant messages that copies a clean Markdown
version of the response by converting Dyad tags and preserving code
blocks. Improves shareability and removes Dyad-only markup; addresses
Linear #1290.
- **New Features**
- Added useCopyToClipboard hook that parses Dyad tags to Markdown,
auto-detects code language, and cleans spacing.
- Updated ChatMessage to show a Copy button (with Copy/Copied tooltip)
to the left of approval status; disabled while streaming.
- Tag conversions: think → "### Thinking"; dyad-write/edit → "###
File/Edit: path" + fenced code; dyad-execute-sql → "### Execute SQL" +
sql block; other Dyad tags map to concise headings; chat-summary/command
are omitted.
- Added e2e tests for clipboard copy, Dyad tag stripping/formatting, and
tooltip states.
<!-- End of auto-generated description by cubic. -->
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Fixes MCP handoff by detecting an edit-code tool call and stopping
generation. Simplifies the title bar credit tooltip; also blocks dyad-*
tags in the system prompt and updates tests.
- **Bug Fixes**
- Stop generation on edit-code via hasToolCall; add a no-op edit-code
tool to signal handoff.
- Combine tool-call stop with existing step limit for reliability.
- Forbid <dyad-*> tags in the system prompt to prevent misuse.
- Remove credit usage/reset details from the title bar tooltip; keep the
delay note.
- Update e2e snapshot to reflect the new prompt rule.
<!-- End of auto-generated description by cubic. -->
This PR implements app search feature and addresses the issue #1182.
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Adds a fast app search with a command-style dialog so users can find
apps by name or chat content and jump to them quickly. Implements the
search experience requested in #1182.
- New Features
- Search dialog (Ctrl+K or “Search Apps” button) with result snippets
from matching chat titles/messages.
- Searches across app names, chat titles, and message content;
case-insensitive; supports partial matches; empty query lists all apps.
- Selecting a result navigates to the app and closes the dialog.
- New IPC endpoint search-app with Zod-validated results, debounced
React Query hook, and preload allowlist update.
- Added E2E tests for dialog open/close, shortcuts, matching behavior,
empty state, and navigation.
<!-- End of auto-generated description by cubic. -->
## 🚀 Feature: Inline Code Editor
This PR adds a comprehensive inline code editing experience to the
DyadWrite component.
### ✨ What's New
- **Inline Monaco Editor**: Edit code directly within the component
using Monaco Editor
- **Cancel/Revert**: Cancel changes and revert to original code state
- **Language Detection**: Automatic syntax highlighting based on file
extensions
- **Theme Support**: Proper dark/light mode theming integration
https://github.com/user-attachments/assets/c44ab622-6b86-403c-904d-3f327f9719e8
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Adds an inline Monaco-based code editor to DyadWrite so users can edit
code blocks in place, then save or cancel changes. Saves stream edits
back to the chat as a dyad-edit block.
- **New Features**
- Inline editor with Edit, Save, and Cancel; preserves original code and
auto-expands when editing.
- Language detection from file extension and dark/light theme support.
- Save streams edits via useStreamChat as <dyad-edit
path="...">...</dyad-edit> tied to the selected chat.
- Non-edit view still uses CodeHighlight; visibility toggle and
in-progress state respected.
- **Refactors**
- ChatMessage now uses DyadMarkdownParser instead of
VanillaMarkdownParser.
<!-- End of auto-generated description by cubic. -->
## Summary
Adds the ability to edit existing custom AI providers through the
settings UI.
## Changes Made
- **UI Changes:**
- Added "Edit Provider" button to custom provider dropdown menu
- Modified `CreateCustomProviderDialog` to support edit mode
- **Backend Changes:**
- Implemented `editCustomLanguageModelProvider` handler in
`language_model_handlers.ts`
- Added corresponding IPC client method
- Database transaction ensures atomicity when updating provider and
associated models
- **Testing:**
- Added comprehensive e2e test covering edit functionality
- Tests verify form pre-population, field updates, and UI persistence
https://github.com/user-attachments/assets/e8c8600e-4fb7-4816-be95-993ede1224d4
## Closes
Fixes#1232 and #1171
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Adds edit support for custom language model providers in Settings. Users
can update provider ID, name, API base URL, and API key env var, with
safe backend updates that also retarget associated models if the ID
changes.
- New Features
- Added “Edit Provider” option in the custom provider menu.
- Dialog supports edit mode with pre-filled fields, unified loading
state, and update button text.
- New IPC handler to edit providers with validation and a transaction;
updates linked models when IDs change.
- IPC client and preload channel updated; React hook exposes
editProvider mutation with cache invalidation.
- Added e2e test covering the full edit flow.
<!-- End of auto-generated description by cubic. -->
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Scoped Azure e2e test env setup to prevent leaking variables across
tests. Added testWithConfigSkipIfWindows and moved Azure env vars into a
per-test preLaunchHook; the Azure test now skips on Windows.
<!-- End of auto-generated description by cubic. -->
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Added OpenRouter as a first-class option in the setup banner and
introduced a reusable provider card component. This streamlines provider
selection and adds E2E coverage for the setup flow.
- **New Features**
- Added SetupProviderCard and used it for Google and OpenRouter in
SetupBanner.
- Clicking Google or OpenRouter routes to the correct provider settings
and logs PostHog events.
- Kept “Other providers” link to Settings.
- Added setup.spec.ts E2E test to verify Google, OpenRouter, and Other
navigation; introduced test config flag showSetupScreen to control the
OPENAI_API_KEY shortcut.
<!-- End of auto-generated description by cubic. -->
Based on https://github.com/dyad-sh/dyad/pull/1116
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Adds a fast chat search dialog (Command Palette) to find and jump
between chats. Open via the sidebar button or Ctrl/Cmd+K, with title and
message text search plus inline snippets.
- New Features
- Command palette using cmdk with keyboard shortcut (Ctrl/Cmd+K).
- Searches within the selected app across chat titles and message
content via a new IPC route (search-chats).
- Debounced queries (150ms) with React Query; results de-duplicated and
sorted by newest.
- Snippet preview with highlighted matches and custom ranking; selecting
a result navigates and closes the dialog.
- Search button added to ChatList; basic e2e tests added (currently
skipped).
- Dependencies
- Added cmdk@1.1.1.
- Bumped @radix-ui/react-dialog to ^1.1.15 and updated Dialog to support
an optional close button.
<!-- End of auto-generated description by cubic. -->
---------
Co-authored-by: Evans Obeng <iamevansobeng@outlook.com>
Co-authored-by: Evans Obeng <60653146+iamevansobeng@users.noreply.github.com>
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Set “balanced” as the default smart context mode. Users now get balanced
when Smart Files Context is enabled and no mode is set; “conservative”
must be explicitly selected.
- **Refactors**
- Default fallback to balanced in UI and engine (proSmartContextOption
undefined -> "balanced").
- ProModeSelector saves "conservative" explicitly; selector reads
undefined as balanced.
- Updated schema and types to allow "balanced" | "conservative".
- Engine payload now includes smart_context_mode with "balanced" by
default; e2e tests and snapshots updated.
- **Migration**
- No action needed. Existing users without an explicit mode will use
balanced by default; selecting conservative persists.
<!-- End of auto-generated description by cubic. -->
1. Show less common AI providers (secondary) in submenu
2. Show $ signs for rough cost guide
3. Show "Pro" for supported AI providers with Pro is enabled
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Improves the Model Picker UX by grouping less-used providers under an
“Other AI providers” submenu and adding clear cost and Pro indicators.
This makes picking models faster and more informative.
- **New Features**
- Grouped secondary providers under “Other AI providers” using a new
provider.secondary flag (Azure marked secondary).
- Added rough cost hints: models can set dollarSigns and the UI shows a
“$” badge accordingly.
- Shows a “Pro” badge on supported cloud providers when Pro is enabled;
added a “Custom” badge for custom providers.
- Extended types: LanguageModelProvider.secondary and
LanguageModel.dollarSigns; populated values across OpenAI, Anthropic,
Google, and OpenRouter.
<!-- End of auto-generated description by cubic. -->
This allows almost all the iframe sandbox tokens
https://developer.mozilla.org/en-US/docs/Web/HTML/Reference/Elements/iframe#sandbox
except notably for allow-top-navigation*
Note: allow-same-origin should be OK because the main window is under
the `file://` origin while the iframe window is served on
`localhost:####`
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Enable sandboxing on the preview iframe to isolate preview content while
preserving needed capabilities (allow-same-origin, scripts, forms,
popups, modals, orientation/pointer lock, presentation, downloads).
Top-level navigation remains disallowed; allow-same-origin is safe
because the app is file:// and the iframe runs on localhost.
<!-- End of auto-generated description by cubic. -->
Fixes#710
This PR implements comprehensive Azure OpenAI integration for Dyad,
enabling users to leverage Azure
OpenAI models through proper environment variable configuration. The
implementation adds Azure as a
supported provider with full integration into the existing language
model architecture, including support
for GPT-5 models. Key features include environment-based
configuration using `AZURE_API_KEY` and `AZURE_RESOURCE_NAME`,
specialized UI components that provide clear
setup instructions and status indicators, and seamless integration with
Dyad's existing provider system.
The Azure provider leverages the @ai-sdk/azure package (v1.3.25) for
compatibility with the current
TypeScript language model interfaces.
The implementation includes robust error handling for missing
configuration, comprehensive test coverage
with 9 new unit tests covering critical functionality like model client
creation and error scenarios, and
an E2E test for the Azure-specific settings UI.
<img width="1510" height="908" alt="Screenshot 2025-08-18 at 9 14 32 PM"
src="https://github.com/user-attachments/assets/04aa99e1-1590-4bb0-86c9-a67b97bc7500"
/>
---------
Co-authored-by: graphite-app[bot] <96075541+graphite-app[bot]@users.noreply.github.com>
Co-authored-by: Will Chen <willchen90@gmail.com>
Fixes#1037
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Enable JSON file support in codebase scanning so common configs and data
(e.g., package.json, vercel.json, translations) are included. Adds .json
to the allowed extensions and removes special-casing for
package.json/vercel.json.
<!-- End of auto-generated description by cubic. -->
Fixes#554#1049
<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Stop mutating package.json when reading files. readFileWithCache now
returns raw content so all fields are preserved (e.g., packageManager).
Fixes#554 and #1049.
- **Bug Fixes**
- Removed package.json "cleaning" logic and the cleanContent helper.
- Return and cache unmodified file content from both fs and
virtualFileSystem.
<!-- End of auto-generated description by cubic. -->
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Parameterized the system prompt and tokenized it in e2e dumps to make
snapshots smaller and stable. No runtime behavior changes; future prompt
edits won’t churn tests.
- **Refactors**
- Exported BUILD_SYSTEM_PREFIX and BUILD_SYSTEM_POSTFIX from
system_prompt.ts.
- Updated test_helper to replace the full prompt with
${BUILD_SYSTEM_PREFIX}/${BUILD_SYSTEM_POSTFIX} tokens in message dumps.
- Regenerated e2e snapshots to use tokens, reducing ~270 lines per
snapshot.
<!-- End of auto-generated description by cubic. -->
<!-- This is an auto-generated description by cubic. -->
## Summary by cubic
Adds Smart Context v3 with selectable modes (Off, Conservative,
Balanced) and surfaces token savings in chat. Also improves token
estimation by counting per-file tokens when Smart Context is enabled.
- **New Features**
- Smart Context selector in Pro settings with three options.
Conservative is the default when enabled without an explicit choice.
- New setting: proSmartContextOption ("balanced"); undefined implies
Conservative.
- Engine now receives enable_smart_files_context and smart_context_mode.
- Chat shows a DyadTokenSavings card when the message contains
token-savings?original-tokens=...&smart-context-tokens=..., with percent
saved and a tooltip for exact tokens.
- Token estimation uses extracted file contents for accuracy when Pro +
Smart Context is on; otherwise falls back to formatted codebase output.
<!-- End of auto-generated description by cubic. -->
- [x] Update chat_stream_handlers
- [x] Update token handlers
- [x] Update HomeChatInput
- [x] update lexical chat input: do not allow referencing same app
(current app, or other already selected apps)
- [x] I don't think smart context will work on this...
- [x] Enter doesn't clear...
Now users can free to edit their custom models by double clicking any
custom models created in each provider.
Before this, they have to delete -> create a new one.
I simply add an edit panel (which looks the same as 'Add Custom Model')
and integrate that process into the "update" button.
There is one more issue that if a user deletes a model that he was using
in chat, then back to chat, that model would still appear (and work)
unless user chooses a new one.
Tried to modify "delete-custom-model" in language_model_handlers.ts by
the logic that if the name of that model matches the latest using one ->
switch to auto (or default) model. Yet I failed, maybe need more
explanation for this :)
TODOs:
- [x] Do restart when checkout / restore if there is a DB
- [x] List all branches (branch id, name, date)
- [x] Allow checking out versions with no DB
- [x] safeguard to never delete main branches
- [x] create app hook for neon template
- [x] weird UX with connector on configure panel
- [x] tiny neon logo in connector
- [x] deploy to vercel
- [x] build forgot password page
- [x] what about email setup
- [x] lots of imgix errors
- [x] edit file - db snapshot
- [x] DYAD_DISABLE_DB_PUSH
- [ ] update portal doc
- [x] switch preview branch to be read-only endpoint
- [x] disable supabase sys prompt if neon is enabled
- [ ] https://payloadcms.com/docs/upload/storage-adapters
- [x] need to use main branch...
Phase 2?
- [x] generate DB migrations
