# AI-Mate

> AI-Mate is an intelligent agent system specification compiler and generator that translates natural language ideas into complete, production-ready, modular multi-agent system specifications, workflows, and prompts.

## Core Value Proposition
- **Natural Interface**: Converse in plain language to decompose complex business and technical goals.
- **Decomposition**: Automatically structures ideas into a main orchestrator agent, custom sub-agents, precise skills, and execution pipelines.
- **Cross-Platform Compatibility**: Compiles agent specifications into ready-to-use, optimized configurations for major platforms like Codex, Claude.md, Antigravity, or generic prompt builders.

## Target Platform Specifications

### 1. OpenAI Codex / API Builders
Generates Codex-ready commands specifying:
- Main orchestrator agent role.
- Sub-agent definitions.
- Parallel-execution-with-synthesis workflows.
- Skill definition commands.

### 2. Anthropic Claude (CLAUDE.md)
Generates standard `CLAUDE.md` workspace profiles:
- Core system architecture specifications.
- Tasks, roles, and sequential workflows.
- Integration instructions for developer workspaces.

### 3. Antigravity Workspace Blueprint
Generates a YAML-styled managers blueprint including:
- High-level system manager agent.
- Independent sub-agents with dedicated roles (e.g., Finance, Marketing, Menu, Location, Operations).
- Artifact constraints (e.g. expected markdown output).

### 4. Generic Builder Specification
Generates a universal system prompt:
- Comprehensive instructions for standard LLM prompt builders.
- Explicit definition of sub-agents, collaboration workflows, and synthesis constraints.

## User Scenarios & Use Cases
- **AI Businesses**: Multi-agent business expansion and startup modeling.
- **Research Systems**: Parallel data compilation, analysis, and synthesis.
- **Learning Assistants**: Curated curricula, pricing, and domain validation.
- **Coding Systems**: Modular multi-agent code scaffolding.
- **Content Pipelines**: Automated ideation, drafting, and proofreading sequences.

## Architecture and Flow
1. **Input**: User enters a high-level goal in natural language.
2. **Decomposition**: AI-Mate recommends sub-agent systems and dedicated skills.
3. **Refinement**: User refines settings and selects platform format.
4. **Compilation & Export**: Download or copy clean generated system instructions.

## Technical Details
- **Website App**: [mate.lensetek.com](https://mate.lensetek.com/)
- **Landing Page**: [mate.lensetek.com](https://mate.lensetek.com/)
- **Privacy Policy**: [Privacy Policy](https://mate.lensetek.com/privacy)
- **Publisher**: Lensetek AI / Lensetek Group
- **System Requirements**: Web Browser (Chrome, Safari, Firefox, Edge).