cli_engineer-1.2.0 is not a library.
CLI Engineer
Agentic CLI for software engineering automation. It features pluggable LLM providers, task interpretation, planning, execution, review and an agentic loop.
Quick Start
Installation
Basic Usage
CLI Engineer provides specialized commands for different types of software engineering tasks:
# Code generation
# Code review and analysis
# Documentation generation
# Security analysis
# Refactoring assistance
UI Options
Dashboard UI (Default) - Interactive real-time interface with metrics and progress tracking:
# Dashboard with live metrics (default behavior)
# Dashboard + File Logging - logs appear in UI AND saved to file
# Creates: cli_engineer_YYYYMMDD_HHMMSS.log with all session details
Simple Text Mode - Traditional command-line output:
# Simple text output (disable dashboard)
# Simple text + File logging
Key Features:
- 🎛️ Dashboard UI: Default experience with real-time metrics, progress bars, and live log display
- 📄 Simple Text: Clean terminal output for scripts or minimal environments
- 📝 File Logging: Verbose mode (
-v) automatically creates timestamped session logs - 🔄 Dual Output: Dashboard mode with verbose shows logs in UI AND saves to file simultaneously
Documentation
For comprehensive guides and detailed information, see the docs/ directory:
- User Guide - Complete installation, configuration, and usage instructions
- Architecture - Technical architecture and design decisions
- API Reference - Detailed API documentation
- Development Guide - Contributing and development setup
- Documentation Structure - Guide to the documentation organization
Command Reference
| Command | Purpose | Output |
|---|---|---|
code |
Generate new code | Source files in current directory |
review |
Analyze existing code | code_review.md with findings |
docs |
Generate documentation | Documentation files in docs/ |
security |
Security analysis | security_report.md with vulnerabilities |
refactor |
Code improvement | Refactored source files |
Developer Setup
Features
- 🤖 4 LLM Providers: OpenAI (Responses API), Anthropic (Claude 4), Google Gemini, Ollama (local)
- 🧠 Real-Time Thinking: Live reasoning traces from Claude 4, Gemini, and local models; reasoning summaries from o1/o3/o4-mini
- 📊 Dashboard UI (Default): Interactive interface with streaming thoughts and cost tracking
- 📝 Smart Buffering: Intelligent chunking at sentence boundaries for smooth reasoning display
- 💰 Accurate Costs: Real-time token usage and cost calculation from streaming events
- 🔄 Agentic Loop: Iterative planning, execution, and review with transparent reasoning
- 🔒 Local Option: Ollama support for privacy-focused, offline LLM inference (no API keys)
- 📁 Smart Artifacts: Context-aware file generation with proper restrictions
- 🔒 Command-Specific Behavior: Different file permissions and outputs per command type
- 📖 Comprehensive Documentation: Auto-generated docs with examples and API references
See the User Guide for detailed setup and advanced usage instructions.