Autobots

Hierarchical multi-cluster coding swarm CLI powered by NVIDIA NIM models.

Autobots orchestrates multiple AI models as a hierarchical swarm to plan, implement, validate, and repair code against your target repositories. It features a complete tool system (Read, Write, Edit, Glob, Grep), interactive REPL mode, granular permissions, context management, hooks, MCP integration, and code review.

Installation

pip install autobot-swarm

Quick Start

# Set your API key
export NVIDIA_API_KEY="your_key_here"

# Navigate to your project
cd /path/to/your/project

# Initialize, plan, and run
autobots init
autobots plan --goal "Add user authentication"
autobots run --supervised

Features

Core Swarm

• 9 Specialized Clusters — Optimus (planning), UltraMagnus (backend), Jazz (frontend), RedAlert (security), Ratchet (repair), Perceptor (retrieval), Bumblebee (media), Ironhide (simulation), Wheeljack (science)
• Phase-Based Execution — Inspect, Implement, Validate, Repair pipeline
• Context Injection — Reads project documentation and injects into model prompts
• Automatic Repair — Self-healing with validation-driven repair loops
• Rollback Support — Snapshots before writes, undo reverts changes

File Operations

• Read — Read files with offset/limit, line numbers, binary/image support
• Write — Atomic writes with directory creation and path sandboxing
• Edit — Targeted string replacement with exact match, replaceAll mode
• Glob — Pattern-based file search (**/*.py)
• Grep — Regex content search across files

Interactive Mode

• REPL — Conversational session with history and streaming
• One-Shot — autobots ask "question" for quick queries
• Piped Input — cat file.py | autobots ask "explain"
• Slash Commands — /help, /clear, /cost, /compact, /model, /exit

Permissions

• Tool-Level Control — Allow/deny/ask per tool call
• Interactive Approval — y/n/a/Esc prompt
• Config Merge — Global + project + environment settings
• Audit Logging — JSON-lines log of all decisions

Context Management

• Token Tracking — Real-time token usage and budget display
• Auto-Compaction — Summarizes conversation at 80% threshold
• CLAUDE.md Hierarchy — Global, project, and subdirectory instructions

Extensions

• Hooks — Pre/post tool execution callbacks
• MCP — Model Context Protocol server integration
• Plugins — Custom extensions via hook system

Code Review

• Git Diff Review — Analyze changes for issues
• Doctor Command — Health checks for API, connectivity, Python, package

CLI Commands

Configuration

Create .autobots.toml in your project root:

[autobots]
model_selection_profile = "balanced"
default_mode = "supervised"
milestone_threshold = 3
temperature = 0.2
max_tokens = 4096

Execution Modes

• Supervised — Manual approval per phase (default)
• Milestone — Approval every N phases
• Autonomous — No approval gates

Requirements

• Python 3.11+
• NVIDIA API Key

Links

• GitHub: https://github.com/DanielDeshmukh/autobots
• Issues: https://github.com/DanielDeshmukh/autobots/issues
• PyPI: https://pypi.org/project/autobot-swarm/

License

MIT