EclipseWatcher: Using Agents and MCP to Find the latest Berserk Chapter Drops
About
MCP Server and Client to give LLM Agents the ability to search the web for news on a favourite manga of mine: Berserk.
Details
- Author
- KabakaWilliam
- Downloads
- 242
- Categories
- Other
Jump to
- Cloudflare-aware scraper using cloudscraper and BeautifulSoup
- MCP tools auto-described for function-calling LLM agents
- Supports get_latest_berserk_chapter() and is_new_berserk_chapter() queries
- One-line integration with Qwen-client via VLLM
- FastMCP-based server for easy agent integration
Clone the repository, start the VLLM server with start_server.sh, then run the code blocks inside qwen_client.ipynb. The server exposes two MCP tools: get_latest_berserk_chapter() and is_new_berserk_chapter(last_seen), which agents can call to answer questions about the newest chapter.
EclipseWatcher: Using Agents and MCP to Find the latest Berserk Chapter Drops
MCP Server and Client to give LLM Agents the ability to search the web for news on a favourite manga of mine: Berserk.> Automated agent + MCP stack that pings you the moment a new Berserk manga chapter hits _readberserk.com_.
<!--
-->

EclipseWatcher scrapes the Berserk chapter list, wraps the logic in a FastMCP server, and lets any LLM agent running on vLLM answer any variation of :
- “What’s the newest chapter of Berserk?”
Features
| | | |---|---| | 🔎 Scraper | Cloudflare-aware HTML grabber built withcloudscraper + BeautifulSoup. |
| ⚙️ MCP API | get_latest_berserk_chapter() and is_new_berserk_chapter(last_seen) tools auto-described for function-calling agents. |
| 🤖 Agent-ready | One-line integration with Qwen-Client; just run the VLLM server and ask away|
---
Getting Started
1. Clone the repository:
git clone https://github.com/yourusername/agent_test.git
2. Start the VLLM server with
start_server.sh and run the code blocks inside qwen_client.ipynb
Requirements
- VLLM - FastMCP - requests - beautifulsoup4 - websockets - asyncio - Python 3.10Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.



