zarz 0.3.4-alpha → 0.3.5-alpha

package/Cargo.toml

@@ -1,30 +0,0 @@
-[package]
-name = "zarzcli"
-version = "0.3.4-ALPHA"
-edition = "2024"
-
-[dependencies]
-anyhow = "1.0.100"
-clap = { version = "4.5.51", features = ["derive"] }
-dialoguer = "0.12"
-reqwest = { version = "0.12.24", features = ["json", "gzip", "brotli", "stream", "rustls-tls"] }
-serde = { version = "1.0.228", features = ["derive"] }
-serde_json = "1.0.145"
-similar = "2.7.0"
-tokio = { version = "1.48.0", features = ["macros", "rt-multi-thread", "fs", "io-util", "process"] }
-async-trait = "0.1.89"
-rustyline = { version = "17.0.2", features = ["custom-bindings"] }
-crossterm = "0.29.0"
-futures = "0.3.31"
-tokio-stream = "0.1.17"
-eventsource-stream = "0.2.3"
-ignore = "0.4.25"
-walkdir = "2.5.0"
-tree-sitter = "0.25.10"
-syn = { version = "2.0.108", features = ["full", "parsing", "visit"] }
-regex = "1.12.2"
-bytes = "1.10.1"
-toml = "0.9.8"
-quote = "1.0.41"
-dirs = "6.0.0"
-chrono = { version = "0.4.39", features = ["serde"] }

package/QUICKSTART.md

@@ -1,326 +0,0 @@
-# ZarzCLI Quick Start Guide
-
-ZarzCLI is an AI-powered coding assistant that works directly from your command line, similar to Claude Code and Codex CLI.
-
-## Installation
-
-1. Make sure you have Rust installed:
-```bash
-curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
-```
-
-2. Build ZarzCLI:
-```bash
-cd ZarzCLI
-cargo build --release
-```
-
-3. Set up your API key:
-```bash
-# For Claude (Anthropic)
-export ANTHROPIC_API_KEY=sk-ant-your-key-here
-
-# Or for OpenAI
-export OPENAI_API_KEY=sk-your-key-here
-```
-
-## Basic Usage
-
-### 1. Interactive Chat (Default)
-
-Just run `zarz` to start an interactive session:
-
-```bash
-./zarz.bat  # Windows
-./zarz.sh   # Linux/Mac
-```
-
-You'll enter a chat session where you can ask questions and edit code:
-
-```
-Zarz Interactive Session
-Type /help for available commands, /quit to exit
-
-> What does this codebase do?
-Assistant: [AI analyzes your project...]
-
-> /edit src/main.rs
-Loaded src/main.rs for editing
-
-> Add better error handling
-Assistant: [AI suggests changes with file blocks]
-File changes detected. Use /diff to review, /apply to apply
-
-> /diff
-[shows unified diff]
-
-> /apply
-Applied changes to src/main.rs
-
-> /quit
-Goodbye!
-```
-
-### 2. One-Shot Questions
-
-Send a single question and get an answer:
-
-```bash
-zarz --message "Explain what this function does" -f src/auth.rs
-```
-
-Or shorter with alias:
-
-```bash
-zarz --msg "Fix the bug in this file" -f src/main.rs
-```
-
-### 3. Switch Models On-the-Fly
-
-Start with one model, switch to another mid-conversation:
-
-```bash
-zarz
-
-> /model claude-haiku-4-5
-Switched to model: claude-haiku-4-5
-Provider: anthropic
-
-> Add docstrings
-[Uses faster, cheaper Haiku model]
-
-> /model gpt-5-codex
-Switched to model: gpt-5-codex
-Provider: openai
-
-> Optimize this algorithm
-[Uses GPT-5 Codex]
-```
-
-## Available Models
-
-### Anthropic Claude
-
-- **claude-sonnet-4-5-20250929** (default) - Best for complex coding ($3/$15 per 1M tokens)
-- **claude-haiku-4-5** - Fast and cheap ($1/$5 per 1M tokens)
-- **claude-opus-4-1** - Most powerful for complex tasks
-
-### OpenAI
-
-- **gpt-5-codex** - Optimized for coding ($1.25/$10 per 1M tokens)
-- **gpt-4o** - Multimodal
-- **gpt-4-turbo** - Fast
-
-## In-Chat Commands
-
-- `/help` - Show all commands and current model
-- `/model <name>` - Switch AI model
-- `/edit <file>` - Load a file for editing
-- `/diff` - Preview pending changes
-- `/apply` - Apply pending changes
-- `/undo` - Discard pending changes
-- `/run <command>` - Execute shell command
-- `/search <symbol>` - Find symbol in codebase
-- `/context <query>` - Find relevant files
-- `/files` - List loaded files
-- `/clear` - Clear conversation history
-- `/quit` - Exit
-
-## Common Workflows
-
-### Workflow 1: Code Review and Fix
-
-```bash
-zarz
-
-> /edit src/server.rs
-> Review this code for security issues
-[AI finds issues]
-
-> Fix the SQL injection vulnerability
-[AI provides fix with file blocks]
-
-> /diff
-[Review changes]
-
-> /apply
-[Changes applied]
-
-> /run cargo test
-[Verify tests pass]
-```
-
-### Workflow 2: Add Feature with Cost Optimization
-
-```bash
-zarz
-
-# Use Sonnet for planning
-> Design an authentication system with JWT
-
-# Switch to Haiku for implementation (cheaper)
-> /model claude-haiku-4-5
-> Implement the JWT authentication based on your design
-
-# Back to Sonnet for complex parts
-> /model claude-sonnet-4-5-20250929
-> Add rate limiting and refresh token logic
-```
-
-### Workflow 3: Multi-File Refactoring
-
-```bash
-zarz
-
-> /edit src/main.rs
-> /edit src/config.rs
-> /edit src/handlers.rs
-
-> Refactor these files to use async/await consistently
-
-> /diff
-[Shows changes across all files]
-
-> /apply
-[Applies all changes]
-
-> /run cargo check
-[Verify code compiles]
-```
-
-## Legacy Commands (Still Supported)
-
-### Ask Mode
-
-```bash
-zarz ask --prompt "What does this do?" src/main.rs
-```
-
-### Rewrite Mode
-
-```bash
-zarz rewrite --instructions "Add error handling" src/lib.rs
-```
-
-### Explicit Chat Mode
-
-```bash
-zarz chat
-```
-
-## Configuration
-
-### Environment Variables
-
-```bash
-# Model selection
-export ZARZ_MODEL=claude-sonnet-4-5-20250929
-
-# Provider selection
-export ZARZ_PROVIDER=anthropic  # or openai
-
-# Anthropic settings
-export ANTHROPIC_API_KEY=sk-ant-...
-export ANTHROPIC_API_URL=https://api.anthropic.com/v1/messages
-export ANTHROPIC_TIMEOUT_SECS=120
-
-# OpenAI settings
-export OPENAI_API_KEY=sk-...
-export OPENAI_API_URL=https://api.openai.com/v1/chat/completions
-export OPENAI_TIMEOUT_SECS=120
-
-# Advanced settings
-export ZARZ_MAX_OUTPUT_TOKENS=4096
-export ZARZ_TEMPERATURE=0.3
-```
-
-### Per-Command Overrides
-
-```bash
-# Use specific model
-zarz --model claude-haiku-4-5
-
-# Use OpenAI instead of default
-zarz --provider openai --model gpt-5-codex
-
-# Custom endpoint
-zarz --endpoint https://custom-api.example.com
-```
-
-## Tips for Best Results
-
-1. **Be Specific**: "Add error handling to the login function" is better than "improve the code"
-
-2. **Use Context**: Load relevant files with `/edit` before asking questions
-
-3. **Iterate**: Don't try to do everything at once. Make small, focused changes
-
-4. **Review Before Applying**: Always `/diff` before `/apply`
-
-5. **Choose the Right Model**:
-   - Sonnet 4.5: Complex refactoring, architecture decisions
-   - Haiku 4.5: Quick fixes, adding docstrings, simple tasks
-   - GPT-5 Codex: Agentic workflows, algorithm optimization
-
-6. **Cost Optimization**: Start with Sonnet for planning, switch to Haiku for implementation
-
-## Troubleshooting
-
-### API Key Not Found
-
-```bash
-Error: Environment variable ANTHROPIC_API_KEY is required
-
-# Fix:
-export ANTHROPIC_API_KEY=sk-ant-your-key
-```
-
-### Model Not Found
-
-```bash
-Error: Unknown model provider for 'typo-model'
-
-# Fix: Use /model without arguments to see available models
-> /model
-```
-
-### Build Errors
-
-```bash
-# Clean and rebuild
-cargo clean
-cargo build --release
-```
-
-## Next Steps
-
-- Read [MODELS.md](MODELS.md) for detailed model information
-- Check [README.md](README.md) for full documentation
-- Report issues: https://github.com/fapzarz/zarzcli/issues
-
-## Quick Reference Card
-
-```
-┌─────────────────────────────────────────────────┐
-│ ZarzCLI Quick Reference                         │
-├─────────────────────────────────────────────────┤
-│ Start Chat:       zarz                          │
-│ One-Shot:         zarz --message "prompt"       │
-│ With Files:       zarz -f file.rs               │
-│                                                 │
-│ In Chat:                                        │
-│   /help          Show all commands              │
-│   /model <name>  Switch AI model                │
-│   /edit <file>   Load file                      │
-│   /diff          Preview changes                │
-│   /apply         Apply changes                  │
-│   /undo          Discard changes                │
-│   /run <cmd>     Execute command                │
-│   /quit          Exit                           │
-│                                                 │
-│ Models:                                         │
-│   claude-sonnet-4-5-20250929 (default)         │
-│   claude-haiku-4-5 (fast/cheap)                 │
-│   gpt-5-codex (OpenAI)                          │
-└─────────────────────────────────────────────────┘
-```

package/src/cli.rs

@@ -1,201 +0,0 @@
-use std::path::PathBuf;
-
-use clap::{Args, Parser, Subcommand, ValueEnum};
-
-#[derive(Debug, Clone, PartialEq, ValueEnum)]
-pub enum Provider {
-    Anthropic,
-    OpenAi,
-    Glm,
-}
-
-impl Provider {
-    pub fn as_str(&self) -> &'static str {
-        match self {
-            Provider::Anthropic => "anthropic",
-            Provider::OpenAi => "openai",
-            Provider::Glm => "glm",
-        }
-    }
-
-    pub fn from_str(name: &str) -> Option<Self> {
-        match name.to_ascii_lowercase().as_str() {
-            "anthropic" => Some(Provider::Anthropic),
-            "openai" => Some(Provider::OpenAi),
-            "glm" => Some(Provider::Glm),
-            _ => None,
-        }
-    }
-
-    #[allow(dead_code)]
-    pub fn from_env_or_default() -> Self {
-        match std::env::var("ZARZ_PROVIDER")
-            .ok()
-            .as_deref()
-            .map(|v| v.to_ascii_lowercase())
-        {
-            Some(ref v) if v == "openai" => Provider::OpenAi,
-            Some(ref v) if v == "anthropic" => Provider::Anthropic,
-            Some(ref v) if v == "glm" => Provider::Glm,
-            _ => Provider::Anthropic,
-        }
-    }
-}
-
-#[derive(Debug, Parser)]
-#[command(
-    name = "zarz",
-    version,
-    about = "ZarzCLI · AI-assisted code refactoring and rewrites",
-    author = "Fapzarz",
-    long_about = "ZarzCLI - Interactive AI coding assistant\n\nUsage:\n  zarz                      Start interactive chat\n  zarz --message \"prompt\"   Send a single prompt and exit\n  zarz ask \"question\"       Ask mode (legacy)\n  zarz chat                 Chat mode (legacy)"
-)]
-pub struct Cli {
-    /// Send a message and exit (like Claude Code)
-    #[arg(long, visible_alias = "msg")]
-    pub message: Option<String>,
-
-    /// Additional files to include as context
-    #[arg(short = 'f', long)]
-    pub files: Vec<PathBuf>,
-
-    #[command(flatten)]
-    pub model_args: CommonModelArgs,
-
-    /// Working directory for the session
-    #[arg(long)]
-    pub directory: Option<PathBuf>,
-
-    #[command(subcommand)]
-    pub command: Option<Commands>,
-}
-
-#[derive(Debug, Subcommand)]
-pub enum Commands {
-    /// Ask the model a question with optional code context.
-    Ask(AskArgs),
-    /// Rewrite one or more files using the model's response.
-    Rewrite(RewriteArgs),
-    /// Start an interactive chat session with AI code assistance.
-    Chat(ChatArgs),
-    /// Configure API keys and settings.
-    Config(ConfigArgs),
-    /// Manage MCP (Model Context Protocol) servers.
-    Mcp(McpArgs),
-}
-
-#[derive(Debug, Args)]
-pub struct CommonModelArgs {
-    /// Target model identifier (e.g. claude-3-5-sonnet-20241022).
-    #[arg(short, long)]
-    pub model: Option<String>,
-    /// Override the default provider.
-    #[arg(long, value_enum)]
-    pub provider: Option<Provider>,
-    /// Override the default API endpoint.
-    #[arg(long)]
-    pub endpoint: Option<String>,
-    /// Optional system prompt override.
-    #[arg(long)]
-    pub system_prompt: Option<String>,
-    /// Timeout in seconds for the request.
-    #[arg(long)]
-    pub timeout: Option<u64>,
-}
-
-#[derive(Debug, Args)]
-pub struct AskArgs {
-    #[command(flatten)]
-    pub model_args: CommonModelArgs,
-    /// Inline prompt text. If omitted, reads from STDIN.
-    #[arg(short, long)]
-    pub prompt: Option<String>,
-    /// Optional file containing additional instructions.
-    #[arg(long)]
-    pub prompt_file: Option<PathBuf>,
-    /// Additional context files to include in the request.
-    #[arg(value_name = "FILE", num_args = 0..)]
-    pub context_files: Vec<PathBuf>,
-}
-
-#[derive(Debug, Args)]
-pub struct RewriteArgs {
-    #[command(flatten)]
-    pub model_args: CommonModelArgs,
-    /// High-level instructions for the rewrite.
-    #[arg(short, long)]
-    pub instructions: Option<String>,
-    /// File containing rewrite instructions.
-    #[arg(long)]
-    pub instructions_file: Option<PathBuf>,
-    /// Apply the changes without confirmation.
-    #[arg(long)]
-    pub yes: bool,
-    /// Preview diff without writing files.
-    #[arg(long)]
-    pub dry_run: bool,
-    /// Target files that will be rewritten.
-    #[arg(value_name = "FILE", num_args = 1..)]
-    pub files: Vec<PathBuf>,
-}
-
-#[derive(Debug, Args)]
-pub struct ChatArgs {
-    #[command(flatten)]
-    pub model_args: CommonModelArgs,
-    /// Working directory for the session (defaults to current directory).
-    #[arg(long)]
-    pub directory: Option<PathBuf>,
-}
-
-#[derive(Debug, Clone, Args)]
-pub struct ConfigArgs {
-    /// Reset configuration and run interactive setup.
-    #[arg(long)]
-    pub reset: bool,
-    /// Show current configuration.
-    #[arg(long)]
-    pub show: bool,
-}
-
-#[derive(Debug, Clone, Args)]
-pub struct McpArgs {
-    #[command(subcommand)]
-    pub command: McpCommands,
-}
-
-#[derive(Debug, Clone, Subcommand)]
-pub enum McpCommands {
-    /// Add a new MCP server
-    Add {
-        /// Server name
-        name: String,
-        /// Server command (for stdio servers)
-        #[arg(long)]
-        command: Option<String>,
-        /// Server arguments
-        #[arg(long, num_args = 0..)]
-        args: Vec<String>,
-        /// Environment variables (KEY=VALUE)
-        #[arg(long = "env")]
-        env_vars: Vec<String>,
-        /// Server URL (for http/sse servers)
-        #[arg(long)]
-        url: Option<String>,
-        /// Transport type: stdio, http, sse
-        #[arg(long, default_value = "stdio")]
-        transport: String,
-    },
-    /// List all configured MCP servers
-    List,
-    /// Get details of a specific MCP server
-    Get {
-        /// Server name
-        name: String,
-    },
-    /// Remove an MCP server
-    Remove {
-        /// Server name
-        name: String,
-    },
-}