# Agentirc

> Agentic IRC chatbot with multi-provider LLMs (OpenAI, xAI, LM Studio), tool use, MCP, and per-user history

- **Type:** MCP server
- **Install:** `agentstack add mcp-h1ddenpr0cess20-agentirc`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [h1ddenpr0cess20](https://agentstack.voostack.com/s/h1ddenpr0cess20)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [h1ddenpr0cess20](https://github.com/h1ddenpr0cess20)
- **Source:** https://github.com/h1ddenpr0cess20/agentirc

## Install

```sh
agentstack add mcp-h1ddenpr0cess20-agentirc
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# agentirc

An AI-powered IRC agent built on a minimal async IRC bot framework. Supports multiple LLM providers (OpenAI, xAI, LM Studio) with per-user conversation history, tool use, and encrypted persistence.

## Table of Contents

- [Quick Start](#quick-start)
- [CLI Options](#cli-options)
- [Commands](#commands)
- [Documentation](#documentation)
- [License](#license)

## Quick Start

```bash
git clone 
cd agentirc
cp .env.example .env
# Edit .env with your IRC server, nick, and at least one API provider
pip install .
agentirc
```

Or with Docker:

```bash
cp .env.example .env
# Edit .env
docker compose up -d
```

> **Requirements:** Python 3.10+, `httpx`, and optionally `cryptography` for encrypted history persistence.

## CLI Options

```
agentirc [options]
```

| Flag | Description |
|---|---|
| `--env-file PATH` | Path to .env file (default: `.env`) |
| `--debug` | Enable debug logging |
| `--host HOST` | IRC server hostname (overrides `IRC_HOST`) |
| `--port PORT` | IRC server port (overrides `IRC_PORT`) |
| `--nick NICK` | Bot nickname (overrides `IRC_NICK`) |
| `--channels CHANS` | Comma-separated channels (overrides `IRC_CHANNELS`) |
| `--tls` | Connect with TLS (overrides `IRC_USE_TLS`) |
| `--model MODEL` | Default model (overrides `DEFAULT_MODEL`) |
| `--init` | Write a starter `.env` to the current directory and exit |
| `--generate-key` | Generate a Fernet encryption key and exit |

CLI flags override their corresponding environment variables.

## Commands

### User Commands

| Command | Aliases | Description |
|---|---|---|
| `!ai ` | `!chat`, `!ask` | Talk to the AI |
| `!persona ` | | Set a persona and reintroduce |
| `!custom ` | | Set a custom system prompt |
| `!reset` | | Reset conversation to defaults |
| `!stock` | | Reset conversation with no system prompt |
| `!mymodel [name]` | | Show or set your model |
| `!location ` | | Set your location for contextual answers |
| `!x  ` | | Talk as another user |

### Admin Commands

| Command | Description |
|---|---|
| `!model [name\|reset]` | Show/set global default model |
| `!tools [on\|off\|toggle\|status]` | Enable/disable tool use |
| `!verbose [on\|off\|toggle]` | Toggle verbose mode |
| `!clear` | Clear all conversation state |
| `!country [on\|off\|status]` | Toggle search country filtering |
| `!join ` | Join a channel |
| `!part [#channel] [reason]` | Leave a channel |

Built-in IRC commands (`!ping`, `!time`, `!help`) are also available.

## Documentation

- [docs/configuration.md](docs/configuration.md) -- IRC and AI provider configuration
- [docs/commands.md](docs/commands.md) -- command registry and decorator API
- [docs/extending.md](docs/extending.md) -- subclassing IRCBot, event hooks
- [docs/ai-agent.md](docs/ai-agent.md) -- AI agent architecture, providers, tools, and history
- [docs/ai-output-disclaimer.md](docs/ai-output-disclaimer.md) -- AI output disclaimer and conditions of use
- [docs/not-a-companion.md](docs/not-a-companion.md) -- project scope and intended use

## License

MIT

## Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [h1ddenpr0cess20](https://github.com/h1ddenpr0cess20)
- **Source:** [h1ddenpr0cess20/agentirc](https://github.com/h1ddenpr0cess20/agentirc)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** yes
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** yes
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/mcp-h1ddenpr0cess20-agentirc
- Seller: https://agentstack.voostack.com/s/h1ddenpr0cess20
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.