⚡ Install  ·  🗂 Domains  ·  📦 Bundles  ·  ▶️ Usage  ·  ❓ FAQ

npx @q28i/aieos --claude

✨ What is AIEOS

One command, then forget it exists. Instead of treating AI as a simple question-answering tool or code generator, AIEOS structures collaboration around Socratic inquiry, preference discovery, and opportunity cost reviews—creating reusable, auditable Decision Contracts.

AIEOS is a model-independent Human Intelligence Amplification Runtime (a Decision OS) designed to increase decision quality under uncertainty while improving the user's independent reasoning. It operates with stable database-backed capability registers, running locally with zero external network overhead or dependency lock-in.

⚡ Install — choose your target

One command installs AIEOS specifications, constitutions, and policies directly into your agent's customization directory:

| Target Platform | Command | Destination Path / File | | --- | --- | --- | | Claude Code | npx @q28i/aieos --claude | ~/.claude/skills/ | | Cursor IDE | npx @q28i/aieos --cursor | Creates .cursorrules in project directory | | Gemini CLI | npx @q28i/aieos --gemini | ~/.gemini/skills/ | | Codex CLI | npx @q28i/aieos --codex | ~/.codex/skills/ | | Antigravity IDE | npx @q28i/aieos --antigravity | ~/.gemini/config/skills/ | | OpenCode | npx @q28i/aieos --opencode | ~/.opencode/skills/ | | Kiro CLI | npx @q28i/aieos --kiro | ~/.kiro/skills/ | | All platforms | npx @q28i/aieos --all | Deploys to all of the above |

Local Project Workspaces

If you want to configure a standalone workspace for project-specific vibe coding, run the interactive wizard, the uninstaller, or the recommendation engine:

npx @q28i/aieos                 # run interactive installer wizard
npx @q28i/aieos discover        # scan repository and get capability recommendations
npx @q28i/aieos uninstall       # clean uninstall of integrations

This scaffolds a local .aieos environment containing:

.aieos/
├── system/                 # Managed ONLY by AIEOS (runtime, logs, adapters)
├── project/                # Project intelligence (memory, decisions, contracts)
├── config/                 # Configuration files (settings.json, tools.json)
├── skills/                 # Custom local capability packages
└── state/                  # Current session state and checkpoints

📦 Themed Bundles

Optionally filter which capabilities get installed onto your agent platform:

npx @q28i/aieos --claude --bundle research     # only research & statistics capabilities
npx @q28i/aieos --cursor --bundle security     # only security capabilities
npx @q28i/aieos --bundle full                  # install everything (default)

🗂 The 4 Domains

AIEOS shifts the execution paradigm from raw task processing to a values-aligned decision pipeline:

| Domain | Scope | Key Abstractions | | --- | --- | --- | | 📜 Constitutions | CONSTITUTION/ | User Agency, Engineering Quality, Security Constraints | | 🔌 Services | SERVICES/ | Kernel Orchestration, SQLite Memory, EventBus, Evolution | | 🧠 Protocols | PROTOCOLS/ | RealityCheck, Curiosity, Wisdom, Judgment Loops | | ⚙️ Policies | POLICIES/ | CognitiveBias Filters, LearningProgress, KnowledgeROI |

📁 Repository Layout

aieos/
├─ 📂 AIEOS/                     # Dynamic system specifications & schemas
├─ 📂 bin/                       # Node.js binary wrappers & environment gates
├─ 📂 doc_system/                # Core Python modules & engines
│   ├─ cli.py                    # CLI installer & route manager
│   ├─ doctor.py                 # Workspace integrity diagnostics
│   ├─ generator.py              # Package scaffolding generator
│   └─ registry.py               # Capability database mappings
├─ 📂 tests/                     # Core integration unit tests
├─ 🗂️ aieos.json                 # Project registry settings
├─ 📜 LICENSE                    # MIT License
└─ 📖 README.md

▶️ Using AIEOS

1. Prompt Socratic Decisions: Once AIEOS rules are active in your agent, prompt it:

   "I want to choose between Serverless SQL and standard Postgres for our analytics workload. Run an AIEOS Decision Protocol."

2. Contract Output: The agent will challenge your assumptions, simulate trade-offs, identify evidence gaps, and emit a structured Decision Contract in your workspace.

📋 System Requirements

• Node.js: 18.0.0 or higher
• Python: 3.11 or higher (automatically verified on startup)
• Git: Required for remote capability package operations

❓ FAQ

🤝 Contributing

PRs are welcome! Feel free to file issues or submit pull requests to extend agent target support or refine cognitive protocols.

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

⬆ back to top