log-experiment
Skill0 starsChanged 19 days agoMITView on GitHub
Close an experiment when the founder says "log NNNN" or shares its result - write the learning card, set the card header to done with the decision, update the canvas box, propose exactly one next hypothesis, sync the dashboard.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- subarnasaikia/kaggle-skillstheir own version, first seen Apr 25, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- subarnasaikia/kaggle-skills★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository6 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- get-startedSkill
- new-experimentSkill
- pull-evidenceSkill
- sync-dashboardSkill
- tourSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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
- Very specific: line 8
Context: Read experiments/NNNN-*/test-card.md (the header, the bar and the How we measure table), and the run-log.md and …
- Very specific: line 10
Final numbers: If the measurement window has not closed, stop: write an interim entry through /pull-evidence NNNN instead and tell the …
- Distinctive: line 31
Next hypothesis: If Pivot or Kill: propose exactly ONE next hypothesis, in the priority order Desirability -> Feasibility -> Viability, …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
Log the result of experiment NNNN (the argument). Honest reading first, decision second.
- Context. Read
experiments/NNNN-*/test-card.md(the header, the bar and the How we measure table), and therun-log.mdanddaily-evidence-log.mdif they exist. - Final numbers. If the measurement window has not closed, stop: write an interim entry
through
/pull-evidence NNNNinstead and tell the founder when the final read is due. A learning card needs the closed window unless the founder explicitly says "log it now". - Learning card. Copy
experiments/_template/learning-card.mdinto the folder and fill it:- Observation: what happened, with the numbers and any verbatim quotes (no personal data)
- Bar: met / not met. People: N, test and internal accounts left out. Per person: matched or not
- Insight: what it means, with the caveats (a weekend in the window, untracked links, small N)
- Decision: Persevere only with a number that clears the bar; otherwise Pivot or Kill
- What happened to the box: new tag and source for the canvas box row
- Header. In the test card's header set
status: done,decision: persevere|pivot|kill,result(the whole read in one sentence), the scoreboard fields/pull-evidencestep 4 lists (read_at,score_now,score_target,score_label,score_parts,score_caveat) for the final read,bar_met: true|false(nullwhen there was no usable read),decided: "YYYY-MM-DD",next(the next experiment's ID, once it exists), clearwaiting_on, and rewritenote(one line: the number against the bar and what comes next). If the learning card has a different file name, setlearning_card. Do not editPROGRESS.md: it is generated. If a board fact is corrected rather than updated, record the correction inBOARD-HISTORY.md. - Canvas. In the canvas the box belongs to (
experiments/canvas/VP-A.mdfor one-letter IDs,experiments/canvas/LC-A.mdfor two-letter IDs) change the box's Tag, Evidence and Source cells, raiseversion, setdate, and add a line under Changes. Do not copy the file. A Lean Canvas block counts as covered only once a row reaches Evidence 2. - Next hypothesis. If Pivot or Kill: propose exactly ONE next hypothesis, in the priority order Desirability -> Feasibility -> Viability, across both canvases (the dashboard's Model tab shows the riskiest box of each). Do not create its folder until the founder says to test it.
- Sync. Run
/sync-dashboard. - Report to the founder: the number against the bar, the decision and why, the one proposed next hypothesis, and what is their call.
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.

