# ATO product brief (markdown)

**Name:** ATO (Agentic Tool Optimization)  
**Site:** https://agentictool.ai  
**One-liner:** The local cockpit where every AI follows your rules — war-rooms, receipts, replay, multi-LLM review.

## Surfaces

1. **Desktop app** (Tauri) — humans create agents, watch live runs, open war-rooms  
2. **CLI** — agents and power users; JSON by default  
3. **MCP server** — harness integration  

## Core capabilities

- Multi-LLM war-rooms (one prompt → N models, closer summary)
- Sticky cross-runtime sessions
- Multi-LLM code review (`ato review`) where later reviewers see earlier findings
- Receipts in local SQLite (tokens, cost, tool calls, files touched)
- Cross-runtime replay + cost delta
- Regression detection joined to a configuration ledger
- Cost recommendations when historical data justifies a model swap
- Skills marketplace, agent wizard, automations, embedded terminal

## Runtimes

Native: Claude Code, Codex, Gemini CLI, OpenClaw, Hermes, Ollama  
Plus 15+ API providers. Subscriptions and API keys are first-class.

## Pricing posture

- Open source (MIT) local app + CLI  
- Pro automation tier: https://agentictool.ai/pro  
- Team workspaces for shared war-rooms (see v2.18 post)

## Content map for LLMs

- https://agentictool.ai/llms.txt  
- https://agentictool.ai/llms-full.txt