From 0607b05e3ba59084c4b4569e413145a537491145 Mon Sep 17 00:00:00 2001
From: latte <latte@hiddenden.cafe>
Date: Sat, 10 Jan 2026 21:49:29 +0100
Subject: [PATCH] added claude

---
 CLAUDE.md | 53 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 53 insertions(+)
 create mode 100644 CLAUDE.md

diff --git a/CLAUDE.md b/CLAUDE.md
new file mode 100644
index 0000000..b8eec6a
--- /dev/null
+++ b/CLAUDE.md
@@ -0,0 +1,53 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Build & Run Commands
+
+```bash
+# Install dependencies
+pip install -r requirements.txt
+
+# Run the bot (requires .env with DISCORD_TOKEN and AI provider key)
+python -m daemon_boyfriend
+
+# Run with Docker
+docker-compose up -d
+
+# Syntax check all Python files
+python -m py_compile src/daemon_boyfriend/**/*.py
+```
+
+## Architecture
+
+This is a Discord bot with AI chat (multi-provider) and SearXNG web search.
+
+### Provider Pattern
+The AI system uses a provider abstraction pattern:
+- `services/providers/base.py` defines `AIProvider` abstract class with `generate()` method
+- `services/providers/openai.py`, `openrouter.py`, `anthropic.py` implement the interface
+- `services/ai_service.py` is the factory that creates the correct provider based on `AI_PROVIDER` env var
+- OpenRouter uses OpenAI's client with a different base URL
+
+### Cog System
+Discord commands are organized as cogs in `cogs/`:
+- `ai_chat.py` - `/chat` command and `@mention` handler (primary user interaction)
+- `search.py` - `/search` and `/image` commands using SearXNG
+- `admin.py` - `/status`, `/help`, `/ping`, `/provider` commands
+
+Cogs are auto-loaded by `bot.py` from the `cogs/` directory.
+
+### Configuration
+All config flows through `config.py` using pydantic-settings. The `settings` singleton is created at module load, so env vars must be set before importing.
+
+### Key Design Decisions
+- `ConversationManager` stores per-user chat history in memory with configurable max length
+- `RateLimiter` tracks per-user rate limits using timestamps in a dict
+- Long AI responses are split via `split_message()` in `ai_chat.py` to respect Discord's 2000 char limit
+- The bot responds to @mentions via `on_message` listener, not just slash commands
+
+## Environment Variables
+
+Required: `DISCORD_TOKEN`, plus one of `OPENAI_API_KEY`, `OPENROUTER_API_KEY`, or `ANTHROPIC_API_KEY` depending on `AI_PROVIDER` setting.
+
+Set `DISCORD_GUILD_ID` for instant slash command sync during development (otherwise global sync takes up to 1 hour).