- Fetch and normalize OpenAI and Anthropic model lists through the proxy service.
- Add searchable multi-model selection with localized UI and persisted adapter mappings.
- Cover endpoint resolution, authentication, pagination, and response parsing.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Added support for checkpointing phases and blob management in the forwarder.
- Introduced new types and methods for handling checkpoint blobs, including queuing and publishing checkpoints.
- Enhanced the projector to build checkpoint projections with content-addressed blobs.
- Implemented tests to ensure proper checkpoint blob synchronization and handling of cancellation scenarios.
- Added support for capturing and decoding Fork Chat traffic, including `ForkBackgroundComposer`, `NotifyConversationClone`, and `UploadConversationBlobs`.
- Updated the README to reflect new features and usage instructions for Fork Chat traffic.
- Modified `.gitignore` to include `proto/extensions-cursor-app/`.
- Refactored `Taskfile.yml` to improve error messages related to Cursor extensions.
- Introduced new tests for decoding functionality in `cursor-proxy-debugger`.
- Refactored response handling in newProtoMessage to utilize agentv1 for GetUsableModelsResponse and GetDefaultModelForCliResponse.
- Added new tests for buildCLIModelDetails and encodeCLIModels to ensure correct model details and encoding.
- Updated buildUsableModelsPayload and buildDefaultModelForCliPayload to leverage buildCLIModelDetails for improved model data structure.
- Updated version number to 0.0.43 across all relevant files.
- Added new features and fixes to release notes, including:
- Fix for command recognition issues.
- Fix for summarize command.
- Fix for MiniMax thinking restriction.
- Support for Fork message and conversation linking.
- Introduction of cursor debugger.
- Added QQ and Telegram group information.
- Introduced a new task for running the Cursor protocol debugging proxy.
- Added a build task for the Cursor protocol debugging proxy.
- Updated documentation to include usage instructions for the debugging tool.
- Integrated a tooltip with a button to open the contributor's GitHub profile in the CursorAccountCard component.
- Updated the layout of the account card for better visual organization.
- Removed routing mode options from the configuration view and adjusted related translations in multiple languages.
- Cleaned up unused routing mode logic in the app state management.
Project sidecar context history into Cursor-compatible JSONL transcripts so previous conversations remain readable when referenced from new chats. Backfill existing sessions on startup and preserve Cursor-managed turn status entries during atomic updates.
Desktop Cursor works through the local proxy because it drives the agent
over BidiAppend/RunSSE plus the already-mocked unary endpoints. The
cursor-agent CLI speaks the same agent protocol but calls additional
unary endpoints that had no local handlers, so every request fell into a
wildcard route and came back as HTTP 404, which the Connect client maps
to '[unimplemented] HTTP 404'.
Three independent breaks, one visible symptom:
1. Startup: ServerConfigService/GetServerConfig (only the AiService
variant was mocked), DashboardService/GetTeamAdminSettingsOrEmptyIfNotInTeam
and DashboardService/ListMarketplaces were missing, so the CLI aborted
during session init.
2. Git workspaces: the CLI resolves the repo path-encryption key from
indexingConfig.default{User,Team}PathEncryptionKey in GetServerConfig
when no IDE-stored repo keys exist. The mock returned no indexingConfig,
so repository identity init failed with 'No encryption key found'.
3. Tool execution: every fs tool executor (Ls/Grep/Glob/Shell) consults
the ignore service, which calls getRepoBlockExcludeGlobs() ->
DashboardService/GetTeamReposOrEmptyIfNotInTeam. The 404 propagated as
the tool result error, so model answers arrived but every tool call
returned '[unimplemented] HTTP 404'. Non-git workspaces skip the
repo-block path, which is why tools only failed inside git repos.
Also close the remaining tolerated 404 noise so a CLI session produces
zero unimplemented responses: model listing (GetUsableModels,
GetDefaultModelForCli, GetDefaultModel), dashboard/plugin housekeeping
(GetGlobalCommands, GetEffectiveUserPlugins, RegisterMarketplaceAndPlugins,
GetCliDownloadUrl) and telemetry (AnalyticsService/SubmitLogs,
AnalyticsService/TrackEvents, OTLP /v1/traces).
Additional logging: PolicyMiddleware now includes the request path in the
per-request log line, which is what made this diagnosable from app.log.
leookun