onefirewall-ai 0.1.2

CLI for the OneFirewall AI Gateway — chat, stream, and build agents from your terminal

onefirewall-ai

CLI for the OneFirewall AI Gateway — chat, stream, and build agents from your terminal.

---

Installation

pip install onefirewall-ai

---

Quick Start

1. Login

onefirewall-ai login

You'll be prompted for:

• Endpoint — your OneFirewall AI Gateway URL (e.g. https://onefirewall.ai)
• API Key — generate one from the Profile page in the dashboard

Credentials are saved to ~/.onefirewall-ai/access.json.

2. Chat

onefirewall-ai What is the capital of France?
onefirewall-ai "Explain Kubernetes in simple terms"
onefirewall-ai Write a Python function to reverse a string

The response streams directly to your terminal in real time.

3. Session Management

By default, every new command creates a unique session so conversation history is maintained separately. You can view past sessions or resume a previous session:

# List all recent chat sessions
onefirewall-ai sessions

# Continue an existing session by ID
onefirewall-ai --session <session-id> "What did we just talk about?"

# Create a custom named session
onefirewall-ai --session my-coding-task "Help me write a Python script"

---

Commands

Command	Description
onefirewall-ai	Show help
onefirewall-ai help	Show help
onefirewall-ai login	Configure endpoint & API key
onefirewall-ai sessions	List all recent chat sessions
onefirewall-ai [--session ID] <message>	Send a message and stream the AI response
onefirewall-ai version	Show version

---

Home Directory

All data is stored in ~/.onefirewall-ai/:

~/.onefirewall-ai/
├── access.json          # Endpoint + API key (private)
├── MEMORY.md            # Persistent facts the agent remembers
├── SOUL.md              # System prompt / agent persona
└── sessions/
    └── <session_id>.md  # Full conversation history for each session

Customise your agent

Edit ~/.onefirewall-ai/SOUL.md to change the AI's persona:

# Soul
You are a concise senior DevOps engineer. Always prefer CLI examples.

Edit ~/.onefirewall-ai/MEMORY.md to give the AI persistent context:

# Memory
- My project uses Python 3.12 and FastAPI.
- Always prefer async/await patterns.

---

API Reference

The CLI uses the OpenAI-compatible endpoint:

POST {endpoint}/api/v1/chat/completions

With stream: true and SSE parsing for real-time output.

---

License

MIT © OneFirewall