alex/g3
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

some fixes

Browse Source
This commit is contained in:
Dhanji Prasanna
2025-10-14 12:44:02 +11:00
parent 5110da0c61
commit bb90cc7826
4 changed files with 101 additions and 45 deletions
Download Patch File Download Diff File Expand all files Collapse all files

66
DESIGN.md
View File

@@ -1,4 +1,4 @@
# G3 General Purpose AI Agent - Design Document
# G3 - AI Coding Agent - Design Document
## Overview
@@ -8,7 +8,7 @@ The agent follows a **tool-first philosophy**: instead of just providing advice,
## Core Principles
1. **Tool-First Philosophy**: Solve problems by actively using tools rather than just describing solutions
1. **Tool-First Philosophy**: Solve problems by actively using tools rather than just providing advice
2. **Modular Architecture**: Clear separation of concerns across multiple Rust crates
3. **Provider Flexibility**: Support multiple LLM providers through a unified interface
4. **Modularity**: Clear separation of concerns
@@ -23,11 +23,11 @@ G3 is organized as a Rust workspace with the following crates:
```
g3/
├── src/main.rs # Main entry point
├── src/main.rs # Main entry point (delegates to g3-cli)
├── crates/
│ ├── g3-cli/ # Command-line interface and TUI
│ ├── g3-core/ # Core agent engine and logic
│ ├── g3-providers/ # LLM provider abstractions
│ ├── g3-cli/ # Command-line interface, TUI, and retro mode
│ ├── g3-core/ # Core agent engine, tools, and streaming logic
│ ├── g3-providers/ # LLM provider abstractions and implementations
│ ├── g3-config/ # Configuration management
│ └── g3-execution/ # Code execution engine
├── logs/ # Session logs (auto-created)
@@ -74,7 +74,7 @@ g3/
- Error handling with automatic retry logic
**Key Features:**
- **Context Window Intelligence**: Automatic monitoring with percentage-based tracking (~80% capacity triggers auto-summarization)
- **Context Window Intelligence**: Automatic monitoring with percentage-based tracking (80% capacity triggers auto-summarization)
- **Tool System**: Built-in tools for file operations (read, write, edit), shell commands, and structured output
- **Streaming Parser**: Real-time parsing of LLM responses with tool call detection and execution
- **Session Management**: Automatic session logging with detailed conversation history and token usage
@@ -86,6 +86,7 @@ g3/
- `write_file`: Create or overwrite files with content
- `str_replace`: Apply unified diffs to files with precise editing
- `final_output`: Signal task completion with detailed summaries
- **Project Management**: Workspace handling, requirements.md processing for autonomous mode
### 2. g3-providers: LLM Provider Abstraction
@@ -97,7 +98,7 @@ g3/
**Supported Providers:**
- **Anthropic**: Claude models via API with native tool calling support
- **Databricks**: Foundation Model APIs with OAuth and token-based authentication
- **Databricks**: Foundation Model APIs with OAuth and token-based authentication (default provider)
- **Embedded**: Local models via llama.cpp with GPU acceleration (Metal/CUDA)
- **Provider Registry**: Dynamic provider management and hot-swapping
@@ -119,7 +120,7 @@ g3/
**Execution Modes:**
- **Single-shot**: Execute one task and exit
- **Interactive**: REPL-style conversation with the agent
- **Interactive**: REPL-style conversation with the agent (default mode)
- **Autonomous**: Coach-player feedback loop for complex projects
- **Retro TUI**: Full-screen terminal interface with real-time updates
@@ -139,11 +140,10 @@ g3/
- Multi-language code execution support
- Error handling and result formatting
**Supported Languages:**
- **Bash/Shell**: Direct command execution with streaming output
- **Python**: Script execution via temporary files
- **JavaScript**: Node.js-based execution
- **Extensible**: Framework for adding additional language support
**Supported Execution:**
- **Bash/Shell**: Direct command execution with streaming output (primary use case)
- **Python**: Script execution via temporary files (legacy support)
- **JavaScript**: Node.js-based execution (legacy support)
**Key Features:**
- **Streaming Output**: Real-time command output display
@@ -161,7 +161,7 @@ g3/
- CLI argument integration
**Configuration Hierarchy:**
1. Default configuration (embedded in code)
1. Default configuration (Databricks provider with OAuth)
2. Configuration files (`~/.config/g3/config.toml`, `./g3.toml`)
3. Environment variables (`G3_*`)
4. CLI arguments (highest priority)
@@ -216,7 +216,7 @@ Advanced autonomous operation with coach-player feedback:
## Provider Comparison
| Feature | Anthropic | Databricks | Embedded |
| Feature | Anthropic | Databricks (Default) | Embedded |
|---------|-----------|------------|----------|
| **Cost** | Pay per token | Pay per token | Free after download |
| **Privacy** | Data sent to API | Data sent to API | Completely local |
@@ -242,7 +242,7 @@ max_tokens = 8192
temperature = 0.1
```
### Enterprise Setup (Databricks)
### Enterprise Setup (Databricks - Default)
```toml
[providers]
default_provider = "databricks"
@@ -314,7 +314,7 @@ g3 --retro --theme dracula
# Full-screen terminal interface
```
## Future Enhancements
## Implementation Details
### Planned Features
- **Plugin System**: Custom tool and provider plugins
@@ -341,10 +341,38 @@ g3 --retro --theme dracula
- **Testing**: Unit tests, integration tests, and property-based testing
### Performance Considerations
- **Async-First**: All I/O operations are asynchronous
- **Async-First**: All I/O operations are asynchronous (Tokio runtime)
- **Streaming**: Real-time response processing where possible
- **Memory Efficiency**: Careful memory management for large contexts
- **Caching**: Strategic caching of expensive operations
- **Profiling**: Regular performance profiling and optimization
This design document reflects the current state of G3 as a mature, production-ready AI coding agent with sophisticated architecture and comprehensive feature set.
## Current Implementation Status
### Fully Implemented
- ✅ **Core Agent Engine**: Complete with streaming, tool execution, and context management
- ✅ **Provider System**: Anthropic, Databricks, and Embedded providers with OAuth support
- ✅ **Tool System**: All 5 core tools (shell, read_file, write_file, str_replace, final_output)
- ✅ **CLI Interface**: Interactive mode, single-shot mode, retro TUI
- ✅ **Autonomous Mode**: Coach-player feedback loop with requirements.md processing
- ✅ **Configuration**: TOML-based config with environment overrides
- ✅ **Error Handling**: Comprehensive retry logic and error classification
- ✅ **Session Logging**: Automatic session tracking and JSON logs
- ✅ **Context Management**: Auto-summarization at 80% capacity
### Architecture Highlights
- **Workspace**: 5 crates with clear separation of concerns
- **Dependencies**: Modern Rust ecosystem (Tokio, Clap, Serde, etc.)
- **Streaming**: Real-time response processing with tool call detection
- **Cross-Platform**: Works on macOS, Linux, and Windows
- **GPU Support**: Metal acceleration for local models on macOS
### Key Files
- `src/main.rs`: 6-line entry point delegating to g3-cli
- `crates/g3-core/src/lib.rs`: 2953 lines - main agent implementation
- `crates/g3-cli/src/lib.rs`: 1354 lines - CLI and interaction modes
- `crates/g3-providers/src/lib.rs`: 144 lines - provider trait and registry
- `crates/g3-config/src/lib.rs`: 265 lines - configuration management
- `crates/g3-execution/src/lib.rs`: 284 lines - code execution engine

1
crates/g3-cli/src/ui_writer_impl.rs
View File

@@ -149,7 +149,6 @@ impl UiWriter for ConsoleUiWriter {
}
fn print_agent_prompt(&self) {
print!(" ");
let _ = io::stdout().flush();
}

31
crates/g3-providers/src/anthropic.rs
View File

@@ -274,22 +274,37 @@ impl AnthropicProvider {
let mut current_tool_calls: Vec<ToolCall> = Vec::new();
let mut partial_tool_json = String::new(); // Accumulate partial JSON for tool calls
let mut accumulated_usage: Option<Usage> = None;
let mut byte_buffer = Vec::new(); // Buffer for incomplete UTF-8 sequences
while let Some(chunk_result) = stream.next().await {
match chunk_result {
Ok(chunk) => {
let chunk_str = match std::str::from_utf8(&chunk) {
Ok(s) => s,
// Append new bytes to our buffer
byte_buffer.extend_from_slice(&chunk);
// Try to convert the entire buffer to UTF-8
let chunk_str = match std::str::from_utf8(&byte_buffer) {
Ok(s) => {
// Successfully converted entire buffer, clear it and use the string
let result = s.to_string();
byte_buffer.clear();
result
}
Err(e) => {
error!("Invalid UTF-8 in stream chunk: {}", e);
let _ = tx
.send(Err(anyhow!("Invalid UTF-8 in stream chunk: {}", e)))
.await;
return accumulated_usage;
// Check if this is an incomplete sequence at the end
let valid_up_to = e.valid_up_to();
if valid_up_to > 0 {
// We have some valid UTF-8, extract it and keep the rest for next iteration
let valid_bytes = byte_buffer.drain(..valid_up_to).collect::<Vec<_>>();
std::str::from_utf8(&valid_bytes).unwrap().to_string()
} else {
// No valid UTF-8 at all, skip this chunk and continue
continue;
}
}
};
buffer.push_str(chunk_str);
buffer.push_str(&chunk_str);
// Process complete lines
while let Some(line_end) = buffer.find('\n') {

48
crates/g3-providers/src/databricks.rs
View File

@@ -299,6 +299,7 @@ impl DatabricksProvider {
std::collections::HashMap::new(); // index -> (id, name, args)
let mut incomplete_data_line = String::new(); // Buffer for incomplete data: lines
let accumulated_usage: Option<Usage> = None;
let mut byte_buffer = Vec::new(); // Buffer for incomplete UTF-8 sequences
while let Some(chunk_result) = stream.next().await {
match chunk_result {
@@ -306,29 +307,42 @@ impl DatabricksProvider {
// Debug: Log raw bytes received
debug!("Raw SSE bytes received: {} bytes", chunk.len());
let chunk_str = match std::str::from_utf8(&chunk) {
// Append new bytes to our buffer
byte_buffer.extend_from_slice(&chunk);
// Try to convert the entire buffer to UTF-8
let chunk_str = match std::str::from_utf8(&byte_buffer) {
Ok(s) => {
// Debug: Log raw string content (truncated for large chunks)
if s.len() > 1000 {
debug!(
"Raw SSE string content (first 500 chars): {:?}...",
&s[..500]
);
} else {
debug!("Raw SSE string content: {:?}", s);
}
s
// Successfully converted entire buffer, clear it and use the string
let result = s.to_string();
byte_buffer.clear();
result
}
Err(e) => {
error!("Invalid UTF-8 in stream chunk: {}", e);
let _ = tx
.send(Err(anyhow!("Invalid UTF-8 in stream chunk: {}", e)))
.await;
return accumulated_usage;
// Check if this is an incomplete sequence at the end
let valid_up_to = e.valid_up_to();
if valid_up_to > 0 {
// We have some valid UTF-8, extract it and keep the rest for next iteration
let valid_bytes = byte_buffer.drain(..valid_up_to).collect::<Vec<_>>();
std::str::from_utf8(&valid_bytes).unwrap().to_string()
} else {
// No valid UTF-8 at all, skip this chunk and continue
continue;
}
}
};
buffer.push_str(chunk_str);
// Debug: Log raw string content (truncated for large chunks)
if chunk_str.len() > 1000 {
debug!(
"Raw SSE string content (first 500 chars): {:?}...",
&chunk_str[..500]
);
} else {
debug!("Raw SSE string content: {:?}", chunk_str);
}
buffer.push_str(&chunk_str);
// Process complete lines, but handle incomplete data: lines specially
while let Some(line_end) = buffer.find('\n') {