agentleFS
erikunhaerikunha/vault-of-loci/.agents/skills/counterfactual/SKILL.md

counterfactual

Skill0 starsChanged 5 days agoMITView on GitHub

Ask the learner what would change if one premise of a concept changed, so the answer is about the mechanism and not a repeated recommendation, then grade it against the source and record the attempt; use it when the learner wants to understand a concept deeper, and it writes a note, so run it only when asked.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 4

    Distinctive

    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
  1. Very specific:

    Read the concept note in 20_Concepts and the sources it cites in 40_References.

    line 13
  2. Distinctive:

    Ask what changes and why. The learner answers first, without notes.

    line 17
  3. Distinctive:

    If the answer only repeats the recommendation, say that the recommendation repeating is not an answer and ask for the mechanism instead of …

    line 18
  4. Distinctive:

    If the answer is wrong, or the learner asks to try again, use the hint ladder one step at a time: hint, strong hint, partial example, …

    line 23
  5. Distinctive:

    Ask for confidence from 1 to 5 before the reveal.

    line 25
  6. Very specific:

    Write the practice note from 90_System/Templates/Counterfactual.md into the practice folder of the track the concept belongs to: …

    line 27

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Tutor mode: the learner attempts first and without notes. The agent never writes the learner's notes or answers. 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.

  1. Take the concept the learner names. If they name none, take the concept of the open note or the weakest one in the record, and say which. Do not ask the learner to choose one.
  2. Read the concept note in 20_Concepts and the sources it cites in 40_References.
  3. Pick one premise from the concept note yourself, in the same reply as step 1: a precondition the idea depends on, a limit, an assumption it rests on. State the counterfactual in one line: "if this premise were not true, what would change?"
  4. Ask what changes and why. The learner answers first, without notes. Wait for the answer.
  5. If the answer only repeats the recommendation, say that the recommendation repeating is not an answer and ask for the mechanism instead of giving the answer: which step breaks, and what follows from that.
  6. Grade the learner's words against the source and the rubric, not against how the conversation went. A fluent answer that never names the mechanism gets no credit.
  7. If the answer is wrong, or the learner asks to try again, use the hint ladder one step at a time: hint, strong hint, partial example, answer. Never skip a step.
  8. Ask for confidence from 1 to 5 before the reveal. Then reveal what the source says the mechanism is.
  9. Write the practice note from 90_System/Templates/Counterfactual.md into the practice folder of the track the concept belongs to: 10_Tracks/<NN>_<Track>/practice/Counterfactual.md. Fill the Date, Concept, Premise changed, What changes and Why columns. Write the note only after the learner confirms it.
  10. Record the attempt after the learner confirms the result. The mode is apply.
python3 -I 90_System/Scripts/record_attempt.py --objective <id> --item <id> --mode apply \
  [--correct|--incorrect] --confidence <1-5> [--hints <n>] [--assistance <level>]
  1. Show the learner the line that was written. Never edit or delete a line of the ledger.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.