experiment-tracking
Skill3 starsChanged 8 months agoMITView on GitHub
Use when running iterative experiments and you need a disciplined way to log configs, seeds, environments, metrics, failures, and comparison summaries.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- kinhluan/skillstheir own version, first seen Nov 8, 2026
- SkillMedev/data-science-mltheir own version, first seen Oct 5, 2026
- gmanch94/claude-code-aidlc-templatetheir own version, first seen Jul 11, 2026
- zhimengj0326/claude-skillstheir own version, first seen Jun 3, 2026
- Hollis36/claude-skilltheir own version, first seen Apr 6, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- kinhluan/skills★ 6same name
- gmanch94/claude-code-aidlc-template★ 2same name
- Hollis36/claude-skill★ 1same name
- SkillMedev/data-science-ml★ 1same name
- zhimengj0326/claude-skills★ 0same 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 repository43 other files it gives its agents.
- agent-codingSkill
- benchmark-designSkill
- citation-managementSkill
- code-reproductionSkill
- code-to-paperSkill
- daskSkill
- exploratory-data-analysisSkill
- get-available-resourcesSkill
- github-release-prepSkill
- latex-postersSkill
- literature-reviewSkill
- markdown-mermaid-writingSkill
- markitdownSkill
- matplotlibSkill
- modalSkill
- networkxSkill
- paper-to-codeSkill
- peer-reviewSkill
- plotlySkill
- polarsSkill
- pytorch-lightningSkill
- pyzoteroSkill
- repo-cleanupSkill
- research-planningSkill
- reviewer-responseSkill
- scholar-evaluationSkill
- scientific-critical-thinkingSkill
- scientific-schematicsSkill
- scientific-slidesSkill
- scientific-visualizationSkill
- scientific-writingSkill
- scikit-learnSkill
- seabornSkill
- shapSkill
- stable-baselines3Skill
- statistical-analysisSkill
- statsmodelsSkill
- sympySkill
- torch-geometricSkill
- transformersSkill
- umap-learnSkill
- vaexSkill
- venue-templatesSkill
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Keep experiment work organized so results can be compared, explained, and reused in papers or reports.
- Distinctive: line 20
run configs, seeds, code version, and environment details
- Distinctive: line 41
never mix incomparable runs into one conclusion table
- Distinctive: line 42
always record code version or commit reference if available
- Distinctive: line 43
treat failed runs as evidence, not noise to be discarded silently
- Distinctive: line 44
keep final summaries tied to raw run records
“Common” is measured: kept by at least 6 in 10 of the 4 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.
Files
Purpose
Keep experiment work organized so results can be compared, explained, and reused in papers or reports.
When To Use
- running multiple model, prompt, or system variants
- comparing baselines and ablations across many runs
- preparing result summaries for a paper or internal update
- diagnosing why numbers changed across commits or environments
Inputs
- experiment goal and benchmark definition
- run configs, seeds, code version, and environment details
- metric outputs, logs, and notes about failures or anomalies
Outputs
- canonical experiment ledger
- comparison table across runs
- best-run summary with supporting evidence
- anomaly log and suspected causes
- next-step recommendation list
Workflow
- Define a stable schema for run IDs, configs, seeds, and artifacts.
- Log every run with enough context to reproduce it later.
- Separate successful, failed, and invalid runs explicitly.
- Compare results only across compatible settings.
- Summarize what changed, what improved, and what still needs testing.
Constraints
- never mix incomparable runs into one conclusion table
- always record code version or commit reference if available
- treat failed runs as evidence, not noise to be discarded silently
- keep final summaries tied to raw run records
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.