coder-config 0.52.2-beta → 0.52.3-beta

package/README.md

@@ -487,7 +487,7 @@ claude plugin install python-support@claude-config-plugins
 1. Open Project Explorer
 2. Click the **+** menu on any project folder
 3. Select **Install Plugins**
-4. Toggle plugins on/off with scope selection (Project/Global/Local)
+4. Toggle plugins on/off with scope selection (Local/User)
 
 ### Plugin Directory
 
@@ -528,19 +528,21 @@ mcp__github__*           # All GitHub MCP tools
 ```
 
 ### Model Selection
-Choose your preferred Claude model (Sonnet 4, Opus 4.5, etc.)
+Choose your preferred Claude model (Opus 4.6, Sonnet 4.6, Haiku 4.5) and effort level (low/medium/high).
 
 ### Behavior
-- Auto-accept edits
-- Verbose mode
-- Enable/disable MCP servers
+- Show thinking summaries
+- Voice dictation
+- Auto memory
+- Enable all project MCP servers
+- Output style, default shell, CLAUDE.md excludes
 
 ## Gemini CLI Settings
 
 The Web UI provides a visual editor for `~/.gemini/settings.json`:
 
 ### Model Selection
-Choose Gemini model (2.5 Pro, 2.5 Flash, etc.) and enable preview features.
+Choose Gemini model (Auto, Gemini 3.1 Pro, 3 Flash, 2.5 Pro, etc.).
 
 ### Display Options
 Configure theme, token count display, diff view, and streaming.
@@ -679,7 +681,7 @@ npm start
 
 ### Testing
 
-The project has comprehensive test coverage with 230 tests across 10 test files:
+The project has comprehensive test coverage with 608 tests across 18 test files:
 
 ```bash
 # Run all tests

package/lib/constants.js

@@ -2,7 +2,7 @@
  * Constants and tool path configurations
  */
 
-const VERSION = '0.52.2-beta';
+const VERSION = '0.52.3-beta';
 
 // Tool-specific path configurations
 const TOOL_PATHS = {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "coder-config",
-  "version": "0.52.2-beta",
+  "version": "0.52.3-beta",
   "description": "Configuration manager for AI coding tools - Claude Code, Gemini CLI, Codex CLI, Antigravity. Manage MCPs, rules, permissions, memory, and workstreams.",
   "author": "regression.io",
   "main": "config-loader.js",

package/ui/dist/assets/{index-OCQSqQs9.js → index-CG-MVxBS.js}

@@ -1733,17 +1733,29 @@ Search GitHub or npm in the Registry view. Click "Add" to pre-fill the JSON with
     `},"configuring-mcps":{title:"Configuring MCPs",content:`
 ## Configuring MCPs
 
-### Required Fields
+### Transport Types
+
+MCPs support multiple transport modes:
+
+| Transport | Field | Description |
+|-----------|-------|-------------|
+| **stdio** | \`command\` + \`args\` | Local process (most common) |
+| **HTTP** | \`url\` | Streamable HTTP endpoint |
+| **SSE** | \`url\` with \`type: "sse"\` | Server-Sent Events (deprecated, use HTTP) |
+| **WebSocket** | \`url\` with \`type: "ws"\` | WebSocket connection |
+
+### Stdio MCP (Local Process)
 
 - **command** - The executable to run (e.g., "npx", "node", "python")
 - **args** - Array of command-line arguments
+- **env** - Environment variables for the MCP
 
-### Optional Fields
+### HTTP/Remote MCP
 
-- **env** - Environment variables for the MCP
-- **cwd** - Working directory
+- **url** - The endpoint URL
+- **type** - Transport type ("http", "sse", or "ws")
 
-### Example Configuration
+### Example Configurations
 
 \`\`\`json
 {
@@ -1755,12 +1767,9 @@ Search GitHub or npm in the Registry view. Click "Add" to pre-fill the JSON with
         "GITHUB_TOKEN": "\${GITHUB_TOKEN}"
       }
     },
-    "postgres": {
-      "command": "npx",
-      "args": ["-y", "@modelcontextprotocol/server-postgres"],
-      "env": {
-        "DATABASE_URL": "\${DATABASE_URL}"
-      }
+    "remote-server": {
+      "type": "http",
+      "url": "https://mcp.example.com/api"
     }
   }
 }
@@ -2076,20 +2085,28 @@ Choose which Claude model to use.
 
 ### Available Models
 
-- **Claude Sonnet 4** - Fast, capable, good for most tasks
-- **Claude Opus 4** - Most capable, best for complex tasks
-- **Claude Haiku** - Fastest, good for simple tasks
+- **Claude Opus 4.6** - Most capable, best for complex tasks
+- **Claude Sonnet 4.6** - Fast output, great balance of speed and capability
+- **Claude Sonnet 4.5** - Previous generation, balanced
+- **Claude Haiku 4.5** - Fastest, good for simple tasks
 
 ### Setting the Model
 
-In **Claude Code** settings, use the Model dropdown to select your preferred model.
+In **Claude Code** settings, use the Model tab to select your preferred model.
+
+### Effort Level
+
+Control reasoning thoroughness:
+- **Low** - Faster, less thorough
+- **Medium** - Balanced (default)
+- **High** - Most thorough reasoning
 
 ### Per-Task Model
 
 Some tasks may benefit from different models:
-- Use Haiku for quick edits
-- Use Opus for complex refactoring
-- Use Sonnet as a balanced default
+- Use Haiku 4.5 for quick edits and simple tasks
+- Use Opus 4.6 for complex refactoring and architecture
+- Use Sonnet 4.6 as a balanced default
     `},behavior:{title:"Behavior Settings",content:`
 ## Behavior Settings
 
@@ -2097,9 +2114,14 @@ Configure how Claude Code behaves.
 
 ### Available Settings
 
-- **Auto-accept edits** - Automatically accept file changes
-- **Verbose mode** - Show more detailed output
-- **Enable MCPs** - Toggle MCP servers on/off
+- **Respect .gitignore** - Honor .gitignore patterns when searching
+- **Show Thinking Summaries** - Display extended thinking summaries
+- **Voice Dictation** - Enable push-to-talk voice input
+- **Auto Memory** - Allow Claude to save notes across sessions
+- **Enable All Project MCP Servers** - Auto-approve project .mcp.json servers
+- **Output Style** - Named output style for responses
+- **Default Shell** - Bash or PowerShell
+- **CLAUDE.md Excludes** - Skip specific CLAUDE.md files in monorepos
 
 ### Configuration
 
@@ -2109,18 +2131,32 @@ These settings are stored in \`~/.claude/settings.json\` and can be edited in th
 
 Hooks allow you to run custom commands at specific points.
 
-### Available Hooks
+### Hook Events (30 total)
+
+Key events include:
+- **SessionStart** / **SessionEnd** - Session lifecycle
+- **PreToolUse** / **PostToolUse** - Before/after tool execution
+- **UserPromptSubmit** - Before processing user input
+- **SubagentStart** / **SubagentStop** - Agent spawning
+- **FileChanged** / **CwdChanged** - File system events
+- **TaskCreated** / **TaskCompleted** - Task tracking
 
-- **preToolUse** - Before a tool is used
-- **postToolUse** - After a tool is used
-- **notification** - When Claude sends a notification
+### Handler Types
+
+- **command** - Shell script (with async, timeout options)
+- **http** - HTTP POST to a URL
+- **prompt** - Single-turn LLM evaluation
+- **agent** - Spawns a subagent
 
 ### Configuration
 
 \`\`\`json
 {
   "hooks": {
-    "postToolUse": "echo 'Tool used'"
+    "PreToolUse": [{
+      "matcher": "Bash",
+      "hooks": [{ "type": "command", "command": "~/.claude/hooks/pre.sh" }]
+    }]
   }
 }
 \`\`\`
@@ -2130,6 +2166,7 @@ Hooks allow you to run custom commands at specific points.
 - Run linters after file edits
 - Log tool usage
 - Custom notifications
+- Workstream context injection
     `}},xW={"gemini-model":{title:"Model Selection",content:`
 ## Model Selection
 
@@ -2139,14 +2176,16 @@ Choose which Gemini model to use.
 
 | Model | Description |
 |-------|-------------|
-| **Gemini 2.5 Pro** | Most capable, best for complex tasks |
-| **Gemini 2.5 Flash** | Fast and efficient, good balance |
-| **Gemini 2.0 Pro** | Previous generation capable model |
-| **Gemini 2.0 Flash** | Previous generation fast model |
+| **Auto** | CLI picks best model per task |
+| **Gemini 3.1 Pro** | Latest, thinking + multimodal tool use |
+| **Gemini 3 Pro** | Thinking + multimodal tools |
+| **Gemini 3 Flash** | Fast with thinking |
+| **Gemini 2.5 Pro** | Previous generation capable |
+| **Gemini 2.5 Flash** | Previous generation fast |
 
-### Preview Features
+### Model Aliases
 
-Enable **Preview Features** to access experimental models like Gemini 2.5 Pro Preview.
+You can also use aliases: \`auto\`, \`pro\`, \`flash\`, \`flash-lite\`
     `},"gemini-display":{title:"Display Options",content:`
 ## Display Options
 
@@ -2156,10 +2195,14 @@ Configure Gemini CLI's appearance and output.
 
 | Setting | Description |
 |---------|-------------|
-| **Theme** | Color theme (Dark, Light, System) |
-| **Show Token Count** | Display token usage in responses |
-| **Show Diff View** | Use diff format for file changes |
-| **Streaming** | Enable real-time response streaming |
+| **Theme** | Color theme for the CLI interface |
+| **Dynamic Window Title** | Update title with status icons |
+| **Show Line Numbers** | Display line numbers in chat |
+| **Show Citations** | Display citations for generated text |
+| **Hide Context Summary** | Hide GEMINI.md and MCP servers above input |
+| **Hide Footer** | Remove footer from the UI |
+| **Screen Reader Mode** | Accessible plain-text output |
+| **Inline Thinking** | Show model thinking (off or full) |
     `},"gemini-general":{title:"General Settings",content:`
 ## General Settings
 
@@ -2170,8 +2213,18 @@ Core Gemini CLI behavior options.
 | Setting | Description |
 |---------|-------------|
 | **Vim Keybindings** | Enable Vim mode in the CLI |
-| **Auto-save** | Automatically save file changes |
-| **Check for Updates** | Check for new Gemini CLI versions |
+| **Auto Update** | Check for new Gemini CLI versions |
+| **Checkpointing** | Enable session recovery support |
+| **Approval Mode** | Default: ask, Auto Edit: auto-approve edits, Plan: read-only |
+
+### Approval Modes
+
+| Mode | Description |
+|------|-------------|
+| **default** | Ask before tool execution |
+| **auto_edit** | Auto-approve file edits |
+| **plan** | Read-only, no edits allowed |
+| **yolo** | Auto-approve everything (use with caution) |
 
 ### Configuration File
 
@@ -2181,26 +2234,44 @@ Settings are stored in \`~/.gemini/settings.json\`:
 {
   "general": {
     "vimMode": false,
-    "autoSave": true,
-    "checkForUpdates": true
+    "defaultApprovalMode": "default",
+    "enableAutoUpdate": true,
+    "checkpointing": { "enabled": false }
   },
   "model": {
-    "default": "gemini-2.5-flash",
-    "previewFeatures": false
+    "name": "auto"
   }
 }
 \`\`\`
     `},"gemini-sandbox":{title:"Sandbox Mode",content:`
 ## Sandbox Mode
 
-Control command execution safety.
+Control command execution safety. Gemini CLI supports 5 sandbox methods.
 
-### Options
+### Sandbox Methods
 
-| Mode | Description |
-|------|-------------|
-| **Enabled** | Commands run in isolated environment |
-| **Disabled** | Commands run directly on system |
+| Method | Platform | Description |
+|--------|----------|-------------|
+| **sandbox-exec** | macOS | Lightweight seatbelt profiles |
+| **Docker** | Cross-platform | Container-based isolation |
+| **Podman** | Cross-platform | Container-based (rootless) |
+| **gVisor (runsc)** | Linux | Strongest isolation, user-space kernel |
+| **LXC** | Linux | Full system container (experimental) |
+
+### Configuration
+
+\`\`\`json
+{
+  "tools": {
+    "sandbox": {
+      "enabled": true,
+      "command": "docker",
+      "networkAccess": false,
+      "allowedPaths": ["/workspace"]
+    }
+  }
+}
+\`\`\`
 
 ### When to Disable
 
@@ -2800,44 +2871,7 @@ Example:
 ### Known Limitation
 
 Antigravity and Gemini CLI share \`~/.gemini/GEMINI.md\` for global instructions. This is a known conflict in Google's tools.
-    `},"tool-differences":{title:"Tool Differences",content:`
-## Tool Differences
-
-### Configuration Formats
-
-All tools use identical JSON format for MCP server definitions:
-
-\`\`\`json
-{
-  "mcpServers": {
-    "server-name": {
-      "command": "npx",
-      "args": ["-y", "@some/server"],
-      "env": { "API_KEY": "..." }
-    }
-  }
-}
-\`\`\`
-
-### Feature Comparison
-
-| Feature | Claude Code | Gemini CLI | Codex CLI | Antigravity |
-|---------|-------------|------------|-----------|-------------|
-| Type | Terminal CLI | Terminal CLI | Terminal CLI | Full IDE |
-| MCP Config | \`.mcp.json\` | \`.gemini/settings.json\` | \`~/.codex/config.toml\` | \`~/.gemini/antigravity/\` |
-| Config Format | JSON | JSON | TOML | JSON |
-| Env Interpolation | Yes (\`\${VAR}\`) | Yes | Yes | No |
-| Skills | \`.claude/commands/\` | \`.gemini/commands/\` | N/A | Unknown |
-| Rules | \`.claude/rules/\` | \`.gemini/rules/\` | N/A | \`.agent/rules/\` |
-| Instructions | \`CLAUDE.md\` | \`GEMINI.md\` | N/A | \`GEMINI.md\` |
-
-### What This Means
-
-- Your MCP registry is shared between all tools
-- When you click "Apply Config", all enabled tools get updated
-- Rules can be synced between Claude Code, Gemini CLI, and Antigravity
-- Gemini CLI and Antigravity share global instructions (known conflict)
-    `},"enabling-tools":{title:"Enabling Tools",content:`
+    `},"tool-differences":{title:"Tool Differences",content:'\n## Tool Differences\n\n### Configuration Formats\n\nAll tools use identical JSON format for MCP server definitions:\n\n```json\n{\n  "mcpServers": {\n    "server-name": {\n      "command": "npx",\n      "args": ["-y", "@some/server"],\n      "env": { "API_KEY": "..." }\n    }\n  }\n}\n```\n\n### Feature Comparison\n\n| Feature | Claude Code | Gemini CLI | Codex CLI | Antigravity |\n|---------|-------------|------------|-----------|-------------|\n| Type | Terminal CLI | Terminal CLI | Terminal TUI | Full IDE |\n| MCP Config | `.mcp.json` | `.gemini/settings.json` | `~/.codex/config.toml` | `~/.gemini/antigravity/` |\n| Config Format | JSON | JSON | TOML | JSON |\n| Env Interpolation | Yes (`${VAR}`) | Yes | Yes | No |\n| Skills | `.claude/skills/` | `.gemini/commands/` | Built-in + SKILL.md | Unknown |\n| Rules | `.claude/rules/` | `.gemini/rules/` | Via AGENTS.md | `.agent/rules/` |\n| Instructions | `CLAUDE.md` | `GEMINI.md` | `AGENTS.md` | `GEMINI.md` |\n| Agents | `.claude/agents/` | `.gemini/agents/` | Built-in multi-agent | N/A |\n\n### What This Means\n\n- Your MCP registry is shared between all tools\n- When you click "Apply Config", all enabled tools get updated\n- Rules can be synced between Claude Code, Gemini CLI, and Antigravity\n- Gemini CLI and Antigravity share global instructions (known conflict)\n    '},"enabling-tools":{title:"Enabling Tools",content:`
 ## Enabling Tools
 
 Control which tools receive configuration updates.

package/ui/dist/index.html

@@ -19,7 +19,7 @@
 
     <!-- PWA Manifest -->
     <link rel="manifest" href="/manifest.json">
-    <script type="module" crossorigin src="/assets/index-OCQSqQs9.js"></script>
+    <script type="module" crossorigin src="/assets/index-CG-MVxBS.js"></script>
     <link rel="stylesheet" crossorigin href="/assets/index-BiF3i3y5.css">
   </head>
   <body>