openclacky 0.7.1 → 0.7.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d00dc4e2243d55c56a3f5df9838517d7588a96137b18348c750e42374d15f7d8
-  data.tar.gz: 0e0fb9b1fce48ac1fb5542e3fb4e5b3744aecb4cd9a51bd15ffa6ada3f892212
+  metadata.gz: 05ce65a24427505768d5edd395ace486c9f7ede88a7abcb42770dce4138a319e
+  data.tar.gz: 400a1efafd35cddf6993ba3298830f13eaa1e3cb4502540191e638cb17b22b66
 SHA512:
-  metadata.gz: ab6df815f041afa4fa586547299a61562e544de3ae2425b2e9c7a7c9256c05e7b0f79dc824553ccd810939bf3cf3d6e1376b32f4d8a92acedb03efedcfa1e552
-  data.tar.gz: a6409dd1ab5456f70ae2f6f576f209d47bb4140956caa14ee18815d91c3c901415219212cc73138e959b43531be9f450938fd8e767e63c1075444c520147fa7e
+  metadata.gz: 8bef68ebcadc6a178a697592e2e92b09d8307d6bd789171a21eedb2b3296b23071fe761caf762dfe6e2e399ed76dcc028d232190ae8ce5652c9387337a9c6a39
+  data.tar.gz: f43e0362780a85166435f572ad42a38654d62f9acc66401e5c229bfd6f27605d75844ee232d3f334fbb8414a0b2d6686b2164a10eb4ebe4ef90b54a021bbf163

data/CHANGELOG.md

@@ -7,6 +7,63 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.7.1] - 2026-02-24
+
+This release brings significant user experience improvements, new interaction modes, and enhanced agent capabilities.
+
+### 🎯 Major Features
+
+**Subagent System**
+- Deploy subagent for parallel task execution
+- Subagent mode with invoke_skill tool and code-explorer skill integration
+- Environment variable support and model type system
+
+**Command Experience**
+- Tab completion for slash commands
+- Ctrl+O toggle expand in diff view
+- JSON mode for structured output
+- Streamlined command selection workflow with improved filtering
+
+**Agent Improvements**
+- Idle compression with auto-trigger (180s timer)
+- Improved interrupt handling for tool execution
+- Preview display for edit and write tools in auto-approve mode
+- Enable preview display in auto-approve mode
+
+**Configuration UI**
+- Auto-save to config modal
+- Improved model management UI
+- Better error handling and validation
+
+### Added
+- Quick start guides in English and Chinese
+- Config example and tests for AgentConfig
+
+### Improved
+- Refactored agent architecture (split agent.rb, moved file locations)
+- Simplified thread management in chat command
+- Dynamic width ratio instead of fixed MAX_CONTENT_WIDTH
+- API error messages with HTML detection and truncation
+- Help command handling
+
+### Changed
+- Removed deprecated Config class (replaced by AgentConfig)
+- Removed confirm_edits permission mode
+- Removed keep_recent_messages configuration
+- Removed default model value
+
+### Fixed
+- Use ToolCallError instead of generic Error in tool registry
+- Handle AgentInterrupted exception during idle compression
+- Handle XML tag contamination in JSON tool parameters
+- Prevent modal flickering on validation failure
+- Update agent client when switching models to prevent stale config
+- Update is_safe_operation to not use removed editing_tool? method
+
+### More
+- Optimize markdown horizontal rule rendering
+- Add debug logging throughout codebase
+
 ## [0.7.0] - 2026-02-06
 
 This is a major release with significant improvements to skill system, conversation memory management, and user experience.

data/README.md

@@ -1,75 +1,66 @@
 # OpenClacky
 
-OpenClacky = Lovable + Supabase
-
-## Features
+[![Build](https://img.shields.io/github/actions/workflow/status/clacky-ai/open-clacky/main.yml?label=build&style=flat-square)](https://github.com/clacky-ai/open-clacky/actions)
+[![Release](https://img.shields.io/gem/v/openclacky?label=release&style=flat-square&color=blue)](https://rubygems.org/gems/openclacky)
+[![Ruby](https://img.shields.io/badge/ruby-%3E%3D%203.1.0-red?style=flat-square)](https://www.ruby-lang.org)
+[![Downloads](https://img.shields.io/gem/dt/openclacky?label=downloads&style=flat-square&color=brightgreen)](https://rubygems.org/gems/openclacky)
+[![License](https://img.shields.io/badge/license-MIT-lightgrey?style=flat-square)](LICENSE.txt)
 
-- 💬 Interactive chat sessions with AI models
-- 🤖 Autonomous AI agent with tool use capabilities
-- 📝 Enhanced input with multi-line support and Unicode (Chinese, etc.)
-- 🖼️ Paste images from clipboard (macOS/Linux)
-- 🚀 Single-message mode for quick queries
-- 🔐 Secure API key management
-- 📝 Multi-turn conversation support
-- 🎨 Colorful terminal output
-- 🌐 OpenAI-compatible API support (OpenAI, Gitee AI, DeepSeek, etc.)
-- 🛠️ Rich built-in tools: file operations, web search, code execution, and more
-- ⚡ Prompt caching support for Claude models (reduces costs up to 90%)
+OpenClacky = Lovable + Supabase
 
-## Installation
+**OpenClacky** is a CLI tool for building full-stack web applications — no technical background required. We spent months crafting a **Rails for AI** full-stack architecture that is fully production-ready, with one-click deployment, isolated dev/production environments, and automatic backups.
 
-### Quick Install (Recommended)
+OpenClacky's goal is to deliver the best balance of **AI quality, AI cost, and AI speed**.
 
-**One-line installation** (auto-detects your system):
+## Quick start
 
 ```bash
-curl -sSL https://raw.githubusercontent.com/clacky-ai/open-clacky/main/scripts/install.sh | bash
+$ openclacky
 ```
 
-This script will:
-- Check your Ruby version
-- Install via Homebrew (macOS) if available
-- Install via RubyGems if Ruby >= 3.1.0 is installed
-- Guide you to install Ruby if needed
+- `/config` — Set your API key, model, and base URL
+- `/new <project-name>` — Create a new project
+- Type your requirements and start building
+
+## Why OpenClacky?
+
+|  | **Claude Code** | **Lovable + Supabase** | **OpenClacky** |
+|---|---|---|---|
+| **Target Users** | Professional developers | Non-technical users | Non-technical users |
+| **Tech Stack** | Any | React + Supabase | Rails (full-stack) |
+| **Full-Stack Integration** | ❌ DIY | ⚠️ Frontend/backend split | ✅ Unified full-stack |
+| **Production-Ready** | ❌ Manual setup | ⚠️ Relies on third-party | ✅ Built-in |
+| **One-Click Deploy** | ❌ | ⚠️ Platform lock-in | ✅ Deploy anywhere |
+| **Dev/Prod Isolation** | ❌ | ❌ | ✅ Automatic |
+| **Automatic Backups** | ❌ | ⚠️ Paid feature | ✅ Built-in |
+| **AI Cost Control** | ❌ Pay per token | ❌ Subscription | ✅ Optimally balanced |
+| **Data Ownership** | ✅ | ❌ Platform-owned | ✅ Fully yours |
+| **Interface** | Terminal | Web UI | Terminal |
 
-### Method 1: Homebrew (macOS/Linux)
+## Features
 
-**Best for macOS users** - Automatically handles Ruby dependencies:
+- [x] `/new <project-name>` — Scaffold a full-stack Rails web app in seconds
+- [x] **Skills system** — Specialized AI workflows for deploy, frontend design, PDF, PPTX, and more
+- [x] **Cost monitoring & compression** — Real-time cost tracking, automatic message compression (up to 90% savings)
+- [x] **One-click deployment** — Ship to production with a single command (with Clacky CDE)
+- [x] **Autonomous AI agent** — Multi-step task execution with undo/redo
+- [x] **Multi-provider support** — OpenAI, Anthropic, DeepSeek, and any OpenAI-compatible API
+- [ ] **Time Machine** — Visual history to rewind and branch any point in your project *(coming soon)*
 
-```bash
-brew tap clacky-ai/openclacky
-brew install openclacky
-```
+## Installation
 
-### Method 2: RubyGems (If you already have Ruby >= 3.1.0)
+### Method 1: One-line Install (Recommended)
 
 ```bash
-gem install openclacky
+/bin/bash -c "$(curl -sSL https://raw.githubusercontent.com/clacky-ai/open-clacky/main/scripts/install.sh)"
 ```
 
-### Method 3: From Source (For Development)
+### Method 2: RubyGems
 
-```bash
-git clone https://github.com/clacky-ai/open-clacky.git
-cd open-clacky
-bundle install
-bin/clacky
-```
-
-### System Requirements
-
-- **Ruby**: >= 3.1.0 (automatically handled by Homebrew)
-- **OS**: macOS, Linux, or Windows (WSL)
-
-### Uninstallation
+**Requirements:** Ruby >= 3.1.0
 
 ```bash
-# Quick uninstall
-curl -sSL https://raw.githubusercontent.com/clacky-ai/open-clacky/main/scripts/uninstall.sh | bash
-
-# Or manually
-brew uninstall openclacky  # If installed via Homebrew
-gem uninstall openclacky   # If installed via gem
+gem install openclacky
 ```
 
 ## Configuration
@@ -77,141 +68,59 @@ gem uninstall openclacky   # If installed via gem
 Before using Clacky, you need to configure your settings:
 
 ```bash
-clacky config set
+$ openclacky
+
+- /config
 ```
 
 You'll be prompted to enter:
 - **API Key**: Your API key from any OpenAI-compatible provider
-- **Model**: Model name (e.g., `gpt-4`, `deepseek-chat`)
-- **Base URL**: OpenAI-compatible API endpoint (e.g., `https://api.openai.com/v1`)
-
-To view your current configuration:
-
-```bash
-clacky config show
-```
+- **Model**: Model name
+- **Base URL**: OpenAI-compatible API endpoint
 
 ## Usage
 
-### AI Agent Mode (Interactive)
-
-Run an autonomous AI agent in interactive mode. The agent can use tools to complete tasks and runs in a continuous loop, allowing you to have multi-turn conversations with tool use capabilities.
-
-```bash
-# Start interactive agent (will prompt for tasks)
-clacky agent
-
-# Start with an initial task, then continue interactively
-clacky agent "Create a README.md file for my project"
-
-# Auto-approve all tool executions
-clacky agent --mode=auto_approve
-
-# Work in a specific project directory
-clacky agent --path /path/to/project
-
-#### Permission Modes
-
-- `auto_approve` - Automatically execute all tools (use with caution)
-- `confirm_safes` - Auto-approve read-only tools, confirm edits
-- `plan_only` - Generate plan without executing
-
-#### Agent Options
+### Scenario 1: Create a new web app
 
 ```bash
---path PATH                    # Project directory (defaults to current directory)
---mode MODE                    # Permission mode
---verbose                      # Show detailed output
+$ openclacky
+> /new my-blog
+# OpenClacky scaffolds a full-stack Rails app in seconds
+# > Add a posts page with title, content, and author fields
+# > Deploy to production
+# > exit
 ```
 
-#### Cost Control & Memory Management
-
-The agent includes intelligent cost control features:
-
-- **Automatic Message Compression**: When conversation history grows beyond 100 messages, the agent automatically compresses older messages into a summary, keeping only the system prompt and the most recent 20 messages. This dramatically reduces token costs for long-running tasks (achieves ~60% compression ratio).
-
-- **Compression Settings**:
-  - `enable_compression`: Enable/disable automatic compression (default: true)
-  - `keep_recent_messages`: Number of recent messages to preserve (default: 20)
-  - Compression triggers at: ~100 messages (keep_recent_messages + 80)
-
-### List Available Tools
-
-View all built-in tools:
+### Scenario 2: Build a feature in an existing project
 
 ```bash
-clacky tools
+$ cd ~/my-project && openclacky
+# > Add user authentication with email and password
+# > Write tests for the auth flow
+# > exit
 ```
 
-#### Built-in Tools
-
-- **todo_manager** - Manage TODO items for task planning and tracking
-- **file_reader** - Read file contents
-- **write** - Create or overwrite files
-- **edit** - Make precise edits to existing files
-- **glob** - Find files by pattern matching
-- **grep** - Search file contents with regex
-- **shell** - Execute shell commands
-- **web_search** - Search the web for information
-- **web_fetch** - Fetch and parse web page content
-
-### Available Commands
+### Scenario 3: Ask questions about your codebase
 
 ```bash
-clacky agent [MESSAGE]    # Run autonomous agent with tool use
-clacky tools              # List available tools
-clacky config set         # Set your API key
-clacky config show        # Show current configuration
-clacky version            # Show clacky version
-clacky help               # Show help information
+$ openclacky
+# > How does the payment module work?
+# > Where is the user session managed?
+# > exit
 ```
 
-## Examples
-
-### Agent Examples
+## Install from Source
 
 ```bash
-# Start interactive agent session
-clacky agent
-# Then type tasks interactively:
-# > Create a TODO.md file with 3 example tasks
-# > Now add more items to the TODO list
-# > exit
-
-# Auto-approve mode for trusted operations
-clacky agent --mode=auto_approve --path ~/my-project
-# > Count all lines of code
-# > Create a summary report
-# > exit
-
-# Using TODO manager for complex tasks
-clacky agent "Implement a new feature with user authentication"
-# Agent will:
-# 1. Use todo_manager to create a task plan
-# 2. Add todos: "Research current auth patterns", "Design auth flow", etc.
-# 3. Complete each todo step by step
-# 4. Mark todos as completed as work progresses
-# > exit
+git clone https://github.com/clacky-ai/open-clacky.git
+cd open-clacky
+bundle install
+bin/clacky
 ```
 
 ## Development
 
-After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment.
-
-To install this gem onto your local machine, run `bundle exec rake install`. To release a new version, update the version number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git commits and the created tag, and push the `.gem` file to [rubygems.org](https://rubygems.org).
-
-### Testing Agent Features
-
-After making changes to agent-related functionality (tools, system prompts, agent logic, etc.), test with this command:
-
-```bash
-# Test agent with a complex multi-step task using auto-approve mode
-echo "Create a simple calculator project with index.html, style.css, and script.js files" | \
-  bin/clacky agent --mode=auto_approve --path=tmp --max-iterations=20
-
-# Expected: Agent should plan tasks (add TODOs), execute them (create files),
-# and track progress (mark TODOs as completed)
-```
+After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/clacky` for an interactive prompt that will allow you to experiment.
 
 ## Contributing
 

data/bin/clarky

@@ -0,0 +1,11 @@
+#!/usr/bin/env ruby
+
+# Display typo correction message
+warn "\nNote: 'clarky' appears to be a typo!"
+warn "The correct command is 'clacky'"
+
+warn "\nRedirecting to the correct command...\n\n"
+
+# Redirect to clacky
+clacky_path = File.expand_path("clacky", __dir__)
+exec(clacky_path, *ARGV)

data/lib/clacky/agent.rb

@@ -655,9 +655,18 @@ module Clacky
       # Append system prompt suffix as user message (for cache reuse)
       if system_prompt_suffix
         messages = subagent.instance_variable_get(:@messages)
+
+        # Build forbidden tools notice if any tools are forbidden
+        forbidden_notice = if forbidden_tools.any?
+          tool_list = forbidden_tools.map { |t| "`#{t}`" }.join(", ")
+          "\n\n[System Notice] The following tools are disabled in this subagent and will be rejected if called: #{tool_list}"
+        else
+          ""
+        end
+
         messages << {
           role: "user",
-          content: "CRITICAL: TASK CONTEXT SWITCH - FORKED SUBAGENT MODE\n\n#{system_prompt_suffix}",
+          content: "CRITICAL: TASK CONTEXT SWITCH - FORKED SUBAGENT MODE\n\nYou are now running as a forked subagent — a temporary, isolated agent spawned by the parent agent to handle a specific task. You run independently and cannot communicate back to the parent mid-task. When you finish (i.e., you stop calling tools and return a final response), your output will be automatically summarized and returned to the parent agent as a result so it can continue.\n\n#{system_prompt_suffix}#{forbidden_notice}",
           system_injected: true,
           subagent_instructions: true
         }

data/lib/clacky/cli.rb

@@ -506,8 +506,6 @@ module Clacky
             handle_time_machine_command(ui_controller, agent, session_manager)
             next
           when "/clear"
-            # Show user input first
-            ui_controller.append_output(display) if display
             sleep 0.1
             # Clear output area
             ui_controller.layout.clear_output
@@ -523,8 +521,6 @@ module Clacky
             ui_controller.stop
             exit(0)
           when "/help"
-            # Show user input first
-            ui_controller.append_output(display) if display
             sleep 0.1
             ui_controller.show_help
             next

data/lib/clacky/default_skills/code-explorer/SKILL.md

@@ -15,27 +15,20 @@ You are now running in a **forked subagent** mode optimized for fast code explor
 ## Your Mission
 Quickly explore and analyze the codebase to answer questions or gather information.
 
-## Your Capabilities
-- Read files: Use `file_reader` to examine source code
-- Search patterns: Use `grep` to find code patterns and text
-- List files: Use `glob` to discover project structure
-- Web research: Use `web_search` and `web_fetch` if needed
-
 ## Your Restrictions
 - NO modifications: You CANNOT use `write` or `edit` tools
-- NO commands: You CANNOT use `safe_shell` to run commands
 - Read-only: Your role is to ANALYZE, not to change
 
-## Workflow
-1. Understand the request: What information is needed?
-2. Explore efficiently: Use glob to map structure, grep to find patterns, file_reader for details
-3. Analyze findings: Draw insights from the code you read
-4. Report clearly: Provide a concise, actionable summary
+## Workflow — follow this order strictly
 
-## Tips for Efficiency
-- Start with `glob` to understand project structure
-- Use `grep` for targeted searches (faster than reading all files)
-- Only use `file_reader` when you need full file content
-- Focus on answering the specific question asked
+1. **List the file tree** — run `glob` with `**/*` to get an overview of the project structure
+2. **Read README.md** — if it exists, read it to understand the project purpose and layout
+3. **Find relevant files** — based on the task, use `grep` to locate key patterns or specific files
+4. **Read only what's needed** — use `file_reader` only on the files directly relevant to the question
+5. **Report clearly** — provide a concise, actionable summary
 
-Remember: You're running on a cheaper model for speed and cost efficiency. Be thorough but concise!
+## Rules
+- Do NOT read files blindly — always have a reason before opening a file
+- Do NOT read every file in a directory — be selective
+- Prefer `grep` over `file_reader` for finding specific patterns
+- Stop as soon as you have enough information to answer the question

data/lib/clacky/default_skills/new/SKILL.md

@@ -36,7 +36,7 @@ cd <project_name>
 
 ### 5. Success Message
 Tell user:
-- ✅ Project created successfully!
+- Project created successfully!
 - Next step: enter project directory to start development
 - Command: `cd <project_name>`
 
@@ -52,4 +52,4 @@ Response:
 1. Creating a new project named "blog"
 2. Cloning template...
 3. Installing dependencies...
-4. ✅ Done! You can now: `cd blog` to start development
+4. Done! You can now: `cd blog` to start development

data/lib/clacky/providers.rb

@@ -0,0 +1,107 @@
+# frozen_string_literal: true
+
+module Clacky
+  # Built-in model provider presets
+  # Provides default configurations for supported AI model providers
+  module Providers
+    # Provider preset definitions
+    # Each preset includes:
+    # - name: Human-readable provider name
+    # - base_url: Default API endpoint
+    # - api: API type (anthropic-messages, openai-responses, openai-completions)
+    # - default_model: Recommended default model
+    PRESETS = {
+      "anthropic" => {
+        "name" => "Anthropic (Claude)",
+        "base_url" => "https://api.anthropic.com",
+        "api" => "anthropic-messages",
+        "default_model" => "claude-sonnet-4-6",
+        "models" => ["claude-opus-4-6", "claude-sonnet-4-6", "claude-haiku-4"]
+      }.freeze,
+
+      "openrouter" => {
+        "name" => "OpenRouter",
+        "base_url" => "https://openrouter.ai/api/v1",
+        "api" => "openai-responses",
+        "default_model" => "anthropic/claude-sonnet-4-5",
+        "models" => []  # Dynamic - fetched from API
+      }.freeze,
+
+      "minimax" => {
+        "name" => "Minimax",
+        "base_url" => "https://api.minimax.chat/v1",
+        "api" => "openai-completions",
+        "default_model" => "MiniMax-Text-01",
+        "models" => ["MiniMax-Text-01", "MiniMax-M2"]
+      }.freeze,
+
+      "kimi" => {
+        "name" => "Kimi (Moonshot)",
+        "base_url" => "https://api.moonshot.cn/v1",
+        "api" => "openai-completions",
+        "default_model" => "kimi-k2.5",
+        "models" => ["kimi-k2.5"]
+      }.freeze
+    }.freeze
+
+    class << self
+      # Check if a provider preset exists
+      # @param provider_id [String] The provider identifier (e.g., "anthropic", "openrouter")
+      # @return [Boolean] True if the preset exists
+      def exists?(provider_id)
+        PRESETS.key?(provider_id)
+      end
+
+      # Get a provider preset by ID
+      # @param provider_id [String] The provider identifier
+      # @return [Hash, nil] The preset configuration or nil if not found
+      def get(provider_id)
+        PRESETS[provider_id]
+      end
+
+      # Get the default model for a provider
+      # @param provider_id [String] The provider identifier
+      # @return [String, nil] The default model name or nil if provider not found
+      def default_model(provider_id)
+        preset = PRESETS[provider_id]
+        preset&.dig("default_model")
+      end
+
+      # Get the base URL for a provider
+      # @param provider_id [String] The provider identifier
+      # @return [String, nil] The base URL or nil if provider not found
+      def base_url(provider_id)
+        preset = PRESETS[provider_id]
+        preset&.dig("base_url")
+      end
+
+      # Get the API type for a provider
+      # @param provider_id [String] The provider identifier
+      # @return [String, nil] The API type or nil if provider not found
+      def api_type(provider_id)
+        preset = PRESETS[provider_id]
+        preset&.dig("api")
+      end
+
+      # List all available provider IDs
+      # @return [Array<String>] List of provider identifiers
+      def provider_ids
+        PRESETS.keys
+      end
+
+      # List all available providers with their names
+      # @return [Array<Array(String, String)>] Array of [id, name] pairs
+      def list
+        PRESETS.map { |id, config| [id, config["name"]] }
+      end
+
+      # Get available models for a provider
+      # @param provider_id [String] The provider identifier
+      # @return [Array<String>] List of model names (empty if dynamic)
+      def models(provider_id)
+        preset = PRESETS[provider_id]
+        preset&.dig("models") || []
+      end
+    end
+  end
+end

data/lib/clacky/ui2/components/modal_component.rb

@@ -154,14 +154,10 @@ module Clacky
 
             # All fields collected - validate if validator provided
             if validator
-              # Show "Testing..." message with debug info
+              # Show "Testing..." message
               testing_row = start_row + @height - 5
               testing_col = start_col + 3
               print "\e[#{testing_row};#{testing_col}H\e[K"
-              
-              # Show debug info about what we're testing
-              debug_info = "Testing: API=#{@values[:api_key]&.[](0..10)}... URL=#{@values[:base_url]} Model=#{@values[:model]}"
-              print @pastel.dim(debug_info)
               print "\e[#{testing_row + 1};#{testing_col}H\e[K"
               print @pastel.cyan("⏳ Testing connection...")
               STDOUT.flush
@@ -171,17 +167,6 @@ module Clacky
               # Clear testing messages
               print "\e[#{testing_row};#{testing_col}H\e[K"
               print "\e[#{testing_row + 1};#{testing_col}H\e[K"
-              
-              if validation_result[:success]
-                # Validation passed - hide cursor and return values
-                print "\e[?25l"
-                return @values
-              else
-                # Validation failed - show error and loop again to let user correct input
-                @error_message = validation_result[:error] || "Validation failed"
-                # Don't clear modal - just loop again to redraw with error message
-                # This prevents the modal from flickering
-              end
             else
               # No validator - return immediately
               print "\e[?25l"