no-ai-checkpoint
Skill0 starsChanged 5 days agoMITView on GitHub
Run a closed-book quiz on concepts the learner last studied more than 14 days ago, optionally for one track, with a predicted score first, then log the score per concept in last_tested and score. Use when the learner asks for a checkpoint or a no-AI test. Writes properties, so only run 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
- misconception-drillSkill
- mistake-reviewSkill
- new-concept-noteSkill
- new-trackSkill
- 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
6 rules. About 0 are common, 3 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Ask the learner to predict the score they expect, 0 to 100, before the first question.
- Distinctive: line 12
After each answer, before any feedback, ask for confidence from 1 (guess) to 5 (certain).
- Distinctive: line 13
If an answer is wrong, use the hint ladder only when the learner asks to try again: hint, strong hint, partial example, answer.
- Very specific: line 14
Add one "explain without notes" round: the learner explains one concept in their own words, and the agent lists only the gaps (the …
- Very specific: line 16
Write only these properties on the concept, after the learner agrees: last_tested: YYYY-MM-DD, score: <0-100>, confidence: <average, 1-5>.
- Very specific: line 19
Run run-checks after writing.
≈ “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 answers first; the agent never shows an answer, a note excerpt or a summary before the attempt, and never writes the learner's study layer or answers.
- Pick concepts: notes in
20_Conceptswhoselast_testedis empty or older than 14 days (the review-queue skill lists them; 14 days is a default, not a research number). For one track, keep the concepts whosetracksproperty links to it. Take at most 5, weakestscorefirst. - Ask the learner to predict the score they expect, 0 to 100, before the first question. Write nothing yet.
- For each concept, ask 3 questions from its flashcards and Key takeaways, at least one that applies the idea to a new situation. One question at a time. The learner works closed book: no notes, no AI help.
- After each answer, before any feedback, ask for confidence from 1 (guess) to 5 (certain). Then reveal and grade against the note.
- If an answer is wrong, use the hint ladder only when the learner asks to try again: hint, strong hint, partial example, answer. Never skip a step.
- Add one "explain without notes" round: the learner explains one concept in their own words, and the agent lists only the gaps (the
teach-the-agentprotocol). - Score per concept is the percentage of questions right on the first attempt, 0 to 100. Show two gaps: confidence against score, and predicted score against actual score. Those gaps are what to study.
- Write only these properties on the concept, after the learner agrees:
last_tested: YYYY-MM-DD,score: <0-100>,confidence: <average, 1-5>. Useobsidian property:setwhen the app runs, else edit the frontmatter. Optionally save the session from90_System/Templates/Checkpoint.md; the learner fills the answer columns. - Offer to add each missed question to
50_Review/Mistakes.mdthrough the mistake-review skill.
Run run-checks after writing. Do not touch ## Study explanation, ## Key takeaways or the flashcards.
Does 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.

