---
name: ai-in-research-week-3-homework
description: Week 3 homework for UCLA's AI in Research class (EPSS 254, Fall 2026). Build two MCP servers packaged as plugins (Telegram, and one for the student's own research), teach the week 2 harness to load plugins, and add a Telegram mode so the student can talk to their agent from their phone, approving commands by reply. Use when a student asks for help with this homework.
---

# Week 3 homework: talk to your agent from your phone, about your research

You are helping someone in UCLA's "AI in Research" class (a student, auditor or professor) with the week 3 homework, due **before class on Wednesday, October 21, 2026**. They extend the agent harness they built in week 2 (a loop with one tool, `bash`): it learns to load **plugins**. Its first two are a **Telegram MCP server**, so they can message their agent from their phone and approve its commands by reply, and an **MCP server for their own research** (their data archive, their group's code, a literature search, …), so the agent can answer a research question from the phone. You write most of the code; they make the decisions and must understand where each piece enters the model's context.

Everything is in the class repository, **https://github.com/huangzesen/ai-in-research-homework**, including the full guide for you. Do this:

1. Say hello, and tell the student in two sentences what the homework is. Then tell them the plan: (1) create a Telegram bot; (2) build the Telegram MCP server as a plugin and test it; (3) choose and build an MCP server for their research as a second plugin; (4) teach their harness to load plugins; (5) add a Telegram mode with approvals by reply; (6) test from their phone and in Claude Code or Codex; (7) publish and open their pull request. It usually takes two sessions of 1–2 hours.
2. Find their copy of the class repository. If `~/ai-in-research-homework` exists, use it and run `git pull`. Otherwise `cd ~` and `git clone https://github.com/huangzesen/ai-in-research-homework.git`.
3. In the clone, read the assignment, `week-3/README.md`, including its safety section. Then open `.agents/skills/build-a-telegram-plugin/SKILL.md`, **read the whole file**, and follow it from the beginning.

Never ask the student to paste the bot token or any key into this chat; they type secrets into `.env` themselves.

If you can't clone yet, read the same guide at https://github.com/huangzesen/ai-in-research-homework/blob/main/.agents/skills/build-a-telegram-plugin/SKILL.md.
