misconception-drill
Skill0 starsChanged 5 days agoMITView on GitHub
Write two items that look alike for a confusion the learner keeps making, one where each idea applies, and log the pair in 50_Review/Misconceptions.md after the learner confirms it; use it when the learner keeps mixing two ideas, and it writes a file, so run it only when asked.
In this repository35 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- llms.txtllms.txt
- adaptive-scaffoldSkill
- calibration-coachSkill
- compression-challengeSkill
- counterfactualSkill
- defuddleSkill
- dependency-checkSkill
- error-recoverySkill
- json-canvasSkill
- memory-palaceSkill
- mistake-reviewSkill
- new-concept-noteSkill
- new-trackSkill
- no-ai-checkpointSkill
- obsidian-basesSkill
- obsidian-cliSkill
- obsidian-markdownSkill
- pretestSkill
- process-paperSkill
- process-webSkill
- quizSkill
- record-attemptSkill
- recovery-after-inactivitySkill
- retention-auditSkill
- review-queueSkill
- run-checksSkill
- schedule-reviewSkill
- spoken-mockSkill
- teach-the-agentSkill
- transfer-labSkill
- vault-searchSkill
- visual-reconstructionSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 13
Run the report and read the first confusion.
- Very specific: line 19
Read the line for that confusion.
- Distinctive: line 22
Read the concept notes of the two ideas and the sources they cite.
- Distinctive: line 24
Write the two items. They look alike on the surface, so the learner cannot tell them by the words.
- Distinctive: line 29
If the learner asks for help, use the hint ladder one step at a time: hint, strong hint, partial example, answer.
- Distinctive: line 31
Ask for confidence from 1 to 5 before the reveal.
- Very specific: line 40
Do not edit concept notes and do not add flashcards here.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Tutor mode: the learner attempts first and before the key is shown. The agent never writes the learner's own answer, never writes the learner's notes and never picks the answer for them. Hints come one step at a time and only after an attempt: hint, strong hint, partial example, answer. Confidence from 1 to 5 is asked before the reveal. The agent never grades an attempt the learner did not confirm.
- Run the report and read the first confusion.
python3 -I 90_System/Scripts/misconception_report.py- Read the line for that confusion. It gives the objective, how many times it was wrong, how many
of those needed help, the items involved and the open mistakes from
50_Review/Mistakes.mdthat name it. - Read the concept notes of the two ideas and the sources they cite. The report says the exercise to build: two items that look alike.
- Write the two items. They look alike on the surface, so the learner cannot tell them by the words. One item is a case where the first idea applies and the other is a case where the second idea applies. Say which idea applies and why is the learner's job, not the agent's.
- Give the learner both items and wait. Do not show the key and do not say which one applies before the learner answers.
- If the learner asks for help, use the hint ladder one step at a time: hint, strong hint, partial example, answer. Never skip a step.
- Ask for confidence from 1 to 5 before the reveal. Then reveal the key and grade the learner's words against the concept notes and the sources.
- Log the pair in
50_Review/Misconceptions.md, under the confusion log, in the format the file describes:YYYY-MM-DD, [[Objective_id|Objective]]: I answered X, but it is Y. I mix up X and Y. Open.Write the entry only after the learner confirms it. The agent never writes the learner's own answer in the entry, so the learner gives the words for it. - Close a confusion only when the learner tells the two apart without help more than once. A
right answer right after the hint ladder does not close it. When it is closed, mark it
closed YYYY-MM-DDin the log. - Do not edit concept notes and do not add flashcards here. Offer the
mistake-reviewskill when the learner wants a corrected card.
python3 -I 90_System/Scripts/misconception_report.py --min-wrong 2
python3 -I 90_System/Scripts/misconception_report.py --jsonDoes it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

