AI Agents · Guided project
Build a Study Planner Agent with Tools — Guided Project
A Python agent that reads your exam timetable and syllabus, calculates the days left and writes a realistic day-by-day revision plan — with a step limit, logs and tests.
- Intermediate
- ~5 hours
- 5 steps
The brief
Turn the agent from the lessons into a tool you will actually use before exams. Your agent gets three tools — exam dates, syllabus units and date arithmetic — and must produce a revision plan that fits the real number of days. You build the loop yourself, test it with a scripted model, then connect a real LLM.
What you will submit: a GitHub repository with the code, a README explaining the design (tools, loop, limits), sample runs and your evaluation results.
Skills you will use: Tool calling · Agent loop · Error handling · Evaluation
Build it
Step by step
Tick each step as you finish it. 0 of 5 done.
-
Step 1Define the tools
Write get_exam_date(subject), list_topics(subject) and days_until(date) with small dictionaries as data. Each returns a short dict or {"error": …}.
-
Step 2Write the loop
Implement run_agent() with a system prompt, max 8 steps, repeated-call detection and a log line for every action.
-
Step 3Test with a scripted model
Create a fake model that returns a fixed sequence of tool calls, and confirm the loop, errors and limits behave correctly.
-
Step 4Connect a real model
Write a call_model() adapter for any LLM API that supports tool calling, reading the API key from an environment variable.
-
Step 5Evaluate and document
Run 10 test goals (including an unknown subject and an exam that already passed), record success, steps and problems, and write the README.
Self-review before you call it done
Tick each point once your project really meets it. These are the things an interviewer will look for.
Check all 5 points, then mark the project as finished.
Next