* feat: add opencode support
* fix: stabilize opencode session startup
* fix: /models
* fix: improveUI for commands
* fix: format commands.js
* feat: load models through provider adapters
Provider model selection had outgrown a single hardcoded service.
The old service mixed shared caching with provider catalogs and CLI lookup details.
That made stale model lists more likely as providers changed on separate schedules.
Move model discovery behind each provider so lookup lives next to the integration.
The shared service now focuses on provider resolution, caching, persistence, and dedupe.
Return cache metadata and add bypassCache because model availability changes outside the app.
The UI and /models command can show freshness and let users force a provider refresh.
Surface model descriptions while keeping fallback catalogs for unavailable CLIs or SDKs.
* feat(models): resolve active session models through provider adapters
The model inventory command was showing a mix of catalog defaults and
composer-local state instead of the model that is actually active for a
real provider session. That made /models, /cost, and /status
misleading once a session had already started, especially for providers
whose effective runtime model can differ from the optimistic model value
held in the UI.
Introduce an explicit getCurrentActiveModel() contract on
IProviderModels so model resolution lives next to each provider's
catalog logic and uses the provider-native source of truth:
- Claude reads the init event from a resumed stream-json run
- Codex reads model from ~/.codex/config.toml
- Cursor reads lastUsedModel from the chat store.db
- OpenCode reads the persisted session model from opencode.db
- Gemini intentionally returns its default because the CLI does not
provide a reliable active-session lookup
Keep the returned shape intentionally minimal ({ model }). The goal is
to expose only what downstream command consumers need and avoid leaking
provider-specific metadata into a shared transport shape that would
create extra UI coupling and future cleanup cost.
Also make command behavior session-aware: when there is no concrete
session id, do not spawn provider processes or inspect provider session
storage just to answer /models, /cost, or /status. In a new-session
view the correct answer is simply the provider default, and doing more
work there adds latency and unnecessary side effects for no user value.
As part of this, centralize two supporting concerns:
- add a shared helper for building the default current-model result from
a provider catalog so fallbacks stay aligned with DEFAULT
- move leaf-directory validation into shared utils so Cursor session
readers and model lookup code enforce the same path-safety rule
Tests were expanded to cover both the new service delegation path and
the sessionless command behavior, while keeping cache-sensitive tests
isolated from persisted host cache state.
Why this change:
- command output should reflect the model actually driving a session
- new-session views should stay fast and side-effect free
- provider-specific active-model lookup should not be scattered across
routes or UI code
- fallback behavior should be explicit, consistent, and limited to the
provider default when no true active model can be resolved
* feat: support session-scoped model overrides
Model selection was acting like a provider-level preference.
That made resumed sessions drift back to a default or request-time model.
Users expect /models changes made inside a conversation to affect that session.
Store explicit session choices in app-owned ~/.cloudcli state.
This avoids editing provider transcripts or native provider config.
Resolve the effective model before launching each provider runtime.
Claude, Cursor, Codex, Gemini, and OpenCode now honor stored resume choices.
Expose a backend active-model change endpoint for existing sessions.
The models modal can now distinguish default changes from session overrides.
It also shows when a selected model will apply on the next response.
For Claude, stop probing active model state by resuming with a dummy prompt.
Read the indexed JSONL transcript from the end instead.
This preserves provider history while honoring /model stdout or model fields.
Add service tests for adapter delegation and resume-model precedence.
The tests keep cache state, override state, and requested fallback separate.
* feat: make command modal more compact
* fix: preserve opencode session creation events
OpenCode emits the real session id asynchronously on its first JSON output. The runner
registered that id from a helper that could not see the spawned process because
the process reference was scoped inside the model-resolution callback. That
ReferenceError was swallowed by the generic JSON parse fallback, so the client
never received session_created. Without that event, a new OpenCode chat stayed
on / and the assistant stream was not attached to the new session view.
Keep the process reference in the outer spawn scope so registration can update
the active-process map and websocket writer as soon as OpenCode announces the
session id. Split JSON parsing from event processing so malformed non-JSON
output can still stream as raw text, while registration or adapter failures are
surfaced as real errors instead of being hidden as assistant content.
Add a fake opencode executable regression test to lock in the expected lifecycle
ordering: session_created must be sent before live assistant messages, and the
same session id must carry through stream_end and complete.
* fix: clarify model refresh and onboarding providers
OpenCode is now a supported chat provider, but first-run onboarding still only offered
Claude, Cursor, Codex, and Gemini. That made OpenCode harder to discover and
forced users to finish setup before finding the provider in settings or chat.
Adding it to onboarding keeps first-run setup aligned with the providers the
application already supports elsewhere.
The model refresh control was also doing too much visual work. In the new chat
model picker, the previous Hard Refresh label looked like the dialog heading,
which made the primary task unclear. Users open that dialog to choose a model;
refreshing catalogs is only a secondary maintenance action for stale cached
provider model lists.
Rename and reposition the refresh affordance so the model picker reads as a
model picker first. The copy now explains why catalogs are cached, when a refresh
is useful, and that the refresh checks every provider. The /models modal gets the
same clarification so both model-selection surfaces describe the cache behavior
consistently.
* fix: format opencode model catalog labels
OpenCode returns provider-prefixed ids directly from the CLI. Passing those ids through as
labels made the model picker hard to scan: users saw values like
anthropic/claude-3-5-sonnet-20241022 or lowercased, hyphen-split text instead
of readable model names.
Keep the exact OpenCode id as the option value because that is what the CLI
expects, but derive a presentation label for the frontend. The formatter is
intentionally generic rather than a catalog of known providers. It handles common
identifier structure such as provider/model, hyphen-delimited words, v-prefixed
versions, adjacent numeric version tokens, and 8-digit date suffixes.
This keeps OpenCode usable as its model list expands across many upstream
providers without requiring code changes for every new provider or model family.
The description keeps the raw provider-prefixed id visible so users can still
confirm the precise model being selected.
* feat: add more fallback models for cursor
* docs: move model catalog out of shared
The model catalog is no longer a frontend/backend runtime contract.
Keeping it under shared made ownership misleading. It implied the catalog was
application code shared by runtime consumers, even though it now only supports
README links and public API documentation.
Move the catalog into public so it lives beside the docs surfaces that need it.
This gives the API docs a stable, served module and gives README readers a
linkable source without suggesting frontend or backend runtime dependency.
Render the API docs model list from the exported provider registry instead of a
hardcoded Claude/Cursor/Codex subset. That keeps Gemini and OpenCode visible and
makes future provider documentation changes flow through one docs-specific file.
Update README links, provider maintenance notes, and package files so published
artifacts include the standalone docs page and model catalog without relying on
the old shared path.
* fix: simplify empty-state model selector
Keep the provider empty state focused on the setup action users need there:
choosing a model.
The refresh control, cache timestamp, and refresh explanation made the dialog feel
like a cache-management surface.
That extra action is out of place in the empty state, where the goal is to start
a chat with the selected provider and model.
Remove the refresh-specific UI from ProviderSelectionEmptyState and drop the
now-unused refresh/cache props from the ChatMessagesPane pass-through.
Refresh behavior remains available in the dedicated command result flow.
WebSocket Module
This module owns the server-side WebSocket gateway used by:
- Chat streaming (
/ws) - Interactive terminal sessions (
/shell) - Plugin WebSocket passthrough (
/plugin-ws/:pluginName)
It is intentionally structured as small services plus a barrel export in index.ts.
Public API
server/modules/websocket/index.ts exports:
createWebSocketServer(server, dependencies)
Creates and wires the sharedwsserver.connectedClientsandWS_OPEN_STATE
Shared chat client registry and open-state constant used by other modules.
Why Dependency Injection Is Used
The module receives runtime-specific functions from server/index.js instead of importing legacy runtime files directly.
Benefits:
- Keeps module boundaries clean (
server/modules/*architecture rule). - Makes each service easier to test in isolation.
- Keeps WebSocket transport concerns separate from provider runtime concerns.
Service Map
| File | Responsibility |
|---|---|
services/websocket-server.service.ts |
Creates WebSocketServer, binds verifyClient, routes connection by pathname |
services/websocket-auth.service.ts |
Authenticates upgrade requests and attaches request.user |
services/chat-websocket.service.ts |
Handles /ws chat protocol and provider command/session control messages |
services/shell-websocket.service.ts |
Handles /shell PTY lifecycle, reconnect buffering, auth URL detection |
services/plugin-websocket-proxy.service.ts |
Bridges client socket to plugin socket |
services/websocket-writer.service.ts |
Adapts raw WebSocket to writer interface (send, setSessionId, getSessionId) |
services/websocket-state.service.ts |
Holds shared chat client set and open-state constant |
High-Level Architecture
flowchart LR
A[HTTP Server] --> B[createWebSocketServer]
B --> C[verifyWebSocketClient]
B --> D{Pathname}
D -->|/ws| E[handleChatConnection]
D -->|/shell| F[handleShellConnection]
D -->|/plugin-ws/:name| G[handlePluginWsProxy]
D -->|other| H[close()]
E --> I[connectedClients Set]
E --> J[WebSocketWriter]
F --> K[ptySessionsMap]
G --> L[Upstream Plugin ws://127.0.0.1:port/ws]
I --> M[projects.service broadcastProgress]
I --> N[sessions-watcher.service projects_updated]
Connection Handshake + Routing
sequenceDiagram
participant Client
participant WSS as WebSocketServer
participant Auth as verifyWebSocketClient
participant Router as connection router
participant Chat as /ws handler
participant Shell as /shell handler
participant Proxy as /plugin-ws handler
Client->>WSS: Upgrade Request
WSS->>Auth: verifyClient(info)
alt Platform mode
Auth->>Auth: authenticateWebSocket(null)
Auth->>Auth: attach request.user
else OSS mode
Auth->>Auth: read token from ?token or Authorization
Auth->>Auth: authenticateWebSocket(token)
Auth->>Auth: attach request.user
end
alt Auth failed
Auth-->>WSS: false (reject handshake)
else Auth ok
Auth-->>WSS: true
WSS->>Router: on("connection", ws, request)
alt pathname == /ws
Router->>Chat: handleChatConnection(ws, request, deps.chat)
else pathname == /shell
Router->>Shell: handleShellConnection(ws, deps.shell)
else pathname startsWith /plugin-ws/
Router->>Proxy: handlePluginWsProxy(ws, pathname, getPluginPort)
else unknown
Router->>Router: ws.close()
end
end
/ws Chat Flow
When a chat socket connects:
- Add socket to
connectedClients. - Build
WebSocketWriter(capturesuserIdfrom authenticated request). - Parse each incoming message with
parseIncomingJsonObject. - Dispatch by
data.type. - On close, remove socket from
connectedClients.
Chat Message Dispatch
flowchart TD
A[Incoming WS message] --> B[parseIncomingJsonObject]
B -->|invalid| C[send {type:error}]
B -->|ok| D{data.type}
D -->|claude-command| E[queryClaudeSDK]
D -->|cursor-command| F[spawnCursor]
D -->|codex-command| G[queryCodex]
D -->|gemini-command| H[spawnGemini]
D -->|cursor-resume| I[spawnCursor resume]
D -->|abort-session| J[abort by provider]
D -->|claude-permission-response| K[resolveToolApproval]
D -->|cursor-abort| L[abortCursorSession]
D -->|check-session-status| M[is*SessionActive + optional reconnectSessionWriter]
D -->|get-pending-permissions| N[getPendingApprovalsForSession]
D -->|get-active-sessions| O[getActive*Sessions]
Chat Notes
abort-sessionreturns a normalizedcompletemessage withaborted: true.check-session-statusreturns{ type: "session-status", isProcessing }.- Claude status checks can reconnect output stream to the new socket via
reconnectSessionWriter.
/shell Terminal Flow
The shell handler manages persistent PTY sessions keyed by:
<projectPath>_<sessionIdOrDefault>[_cmd_<hash>]
This enables reconnect behavior and isolates command-specific plain-shell sessions.
Shell Lifecycle
stateDiagram-v2
[*] --> WaitingInit
WaitingInit --> ValidateInit: message.type == init
ValidateInit --> ReconnectExisting: session key exists and not login reset
ValidateInit --> SpawnNewPTY: valid path + valid sessionId
ValidateInit --> EmitError: invalid payload/path/sessionId
ReconnectExisting --> Running: attach ws, replay buffer
SpawnNewPTY --> Running: pty.spawn + wire onData/onExit
Running --> Running: input -> pty.write
Running --> Running: resize -> pty.resize
Running --> Running: onData -> buffer + output + auth_url detection
Running --> Exited: onExit
Running --> Detached: ws close
Detached --> Running: reconnect before timeout
Detached --> Killed: timeout reached -> pty.kill
Exited --> [*]
Killed --> [*]
EmitError --> WaitingInit
Shell Behaviors in Detail
init: ReadsprojectPath,sessionId,provider,hasSession,initialCommand,isPlainShell.- Login reset: For login-like commands, existing keyed PTY session is killed and recreated.
- Validation:
Path must exist and be a directory;
sessionIdmust match safe pattern. - Command build: Provider-specific command construction with resume semantics.
- PTY output buffering: Stores up to 5000 chunks for replay on reconnect.
- URL detection:
Strips ANSI, accumulates text buffer, extracts URLs, emits
auth_urlonce per normalized URL, supportsautoOpen. - Close behavior: Socket disconnect does not instantly kill PTY; session is kept alive and terminated on timeout.
/plugin-ws/:pluginName Proxy Flow
sequenceDiagram
participant Client
participant Proxy as handlePluginWsProxy
participant PM as getPluginPort
participant Upstream as Plugin WS
Client->>Proxy: Connect /plugin-ws/:name
Proxy->>Proxy: Validate pluginName regex
alt Invalid name
Proxy-->>Client: close(4400, "Invalid plugin name")
else Valid
Proxy->>PM: getPluginPort(name)
alt Plugin not running
Proxy-->>Client: close(4404, "Plugin not running")
else Port found
Proxy->>Upstream: new WebSocket(ws://127.0.0.1:port/ws)
Client-->>Upstream: relay messages bidirectionally
Upstream-->>Client: relay messages bidirectionally
Upstream-->>Client: close propagation
Client-->>Upstream: close propagation
Upstream-->>Client: close(4502, "Upstream error") on upstream error
end
end
Shared Client Registry and Broadcasts
Only chat sockets (/ws) are tracked in connectedClients.
That shared set is consumed by:
modules/projects/services/projects-with-sessions-fetch.service.tsBroadcastsloading_progresswhile project snapshots are being built.modules/providers/services/sessions-watcher.service.tsBroadcastsprojects_updatedwhen provider session artifacts change.
This design centralizes cross-module realtime fanout without requiring route-local references to WebSocket internals.
Writer Adapter (WebSocketWriter)
WebSocketWriter normalizes chat transport behavior to match existing writer-style interfaces used elsewhere.
Methods:
send(data)
JSON-serializes and sends only if socket is open.setSessionId(sessionId)/getSessionId()
Supports provider session bookkeeping and resume flows.updateWebSocket(newRawWs)
Allows active session stream redirection on reconnect.
Error Handling and Close Codes
Current explicit close codes in this module:
4400: Invalid plugin name4404: Plugin not running4502: Upstream plugin WebSocket error
Other errors:
- Chat handler catches and emits
{ type: "error", error }. - Shell handler catches and writes terminal-visible error output.
- Unknown websocket paths are closed immediately.
Extending This Module
To add a new websocket route:
- Add a new handler service under
services/. - Extend
WebSocketServerDependenciesinwebsocket-server.service.tsif needed. - Add a new pathname branch in the router.
- Wire dependency injection from
server/index.js. - Keep
index.tsas barrel-only export surface.