Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for computer science students

Bug Hunt Debug Journal Agent

The root cause of each bug found and recorded, so repeated mistakes become visible

LearningCodingIT and development
See how it worksGet access
Bug Hunt Debug Journal Agent: what goes in, what the agent does and what you get

What it does

When a program fails, many students change code until it passes and learn nothing. This agent reproduces the error first, so both sides see the same failure. It then proposes a hypothesis about the cause, runs a targeted test or adds a print statement to check it, and reads the result. If the hypothesis fails, it forms the next one from what it just learned and repeats. Once the cause is confirmed, it writes a journal entry with the bug type, the symptom, the cause and the fix idea, and compares it with past entries to flag repeat patterns, such as off-by-one errors in loops. The student applies the fix. Edge case: when the failure is intermittent, the agent runs the test ten times before trusting any result.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Student reports a failing program 2 USES A TOOL Run the program and reproduce the error 3 CHECKS THE RESULT Does the error appear every time? If not: run it 10 times, record the pattern, and use itto guide the hypothesis. Back to step 2. 4 DOES Propose a hypothesis about the cause 5 USES A TOOL Run a targeted test or add a print line to check it 6 CHECKS THE RESULT Did the result support the hypothesis? If not: form a new hypothesis from the result and testagain. Back to step 4. 7 USES A TOOL Apply the proposed fix in a copy and rerun the tests 8 DOES Write the journal entry with bug type, cause and fix 9 DOES Compare with past entries and flag repeat patterns 10 YOU APPROVE Student applies the fix in their own code 11 RESULT Updated bug journal
Read the steps as a list
  1. Student reports a failing program
  2. Run the program and reproduce the error
  3. Does the error appear every time?If not: run it 10 times, record the pattern, and use it to guide the hypothesis. Back to step 2.
  4. Propose a hypothesis about the cause
  5. Run a targeted test or add a print line to check it
  6. Did the result support the hypothesis?If not: form a new hypothesis from the result and test again. Back to step 4.
  7. Apply the proposed fix in a copy and rerun the tests
  8. Write the journal entry with bug type, cause and fix
  9. Compare with past entries and flag repeat patterns
  10. Student applies the fix in their own codeThe agent waits here for your OK.
  11. Updated bug journal

How it decides

A hypothesis is accepted only when a targeted test confirms it and the failure goes away in a copy with the fix applied.

  • Rerun an intermittent failure at least 10 times
  • Stop after 5 failed hypotheses and ask the student for more context
  • Flag a pattern when the same bug type appears 3 times
  • Test fixes only in a copy of the code

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Run count for intermittent bugs (default 10)
  • Hypotheses before asking for help (default 5)
  • Journal format and location
  • Number of repeats before flagging a pattern (default 3)

What keeps you in control

It always asks you first

  • Applying the fix in the real code, done by the student

Hard limits

  • Never edits the student's real code
  • Never marks a bug resolved until the tests pass

It stops when

  • Done: cause confirmed and journal entry written
  • Stop: the error cannot be reproduced after 10 runs

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA student's linked list program crashed on the third delete. The agent reproduced it, guessed a null pointer on an empty list, and printed the head after each call. The head was fine, so the guess failed. The next guess was that the tail pointer went stale. A print confirmed it. The agent logged 'stale pointer after delete' and noted it was the student's third pointer bug that month.

More agents for computer science students

Planning · Learning

Skill Gap Learning Plan Agent

A four-week plan that closes the one or two skill gaps that matter most for the target role

10 steps1 check3 jobs
Learning · Human resources

Mock Interview Practice Loop Agent

Core interview questions answered at or above the bar, with approved best answers saved

9 steps2 checks3 jobs
Coding · Quality control

Test Coverage Gap Closing Agent

Add tests to the riskiest untested code and prove that they catch real bugs

10 steps2 checks11 jobs
Learning · IT and development

Lecture Gap Finder Agent

Find the gaps in understanding early and close them with checked practice

11 steps2 checks3 jobs