mit einem Klick
experiment-advisor
// Given a research goal, use existing skills, training reports, and references to propose a small, concrete experiment plan.
// Given a research goal, use existing skills, training reports, and references to propose a small, concrete experiment plan.
Process CSV data files by cleaning, transforming, and analyzing them. Use this when users need to work with CSV files, clean data, or perform basic data analysis tasks.
分析Excel/CSV字段结构,AI自动生成中英文映射,验证翻译质量,输出统计报告。用于电子表格分析、数据字典创建、字段映射场景。
Deploy a Vibes app to exe.dev VM hosting. Uses nginx on persistent VMs with SSH automation. Supports client-side multi-tenancy via subdomain-based Fireproof database isolation.
Integrate Trackio for experiment tracking in Kaggle competitions. Use PROACTIVELY when user trains models, logs metrics, or manages experiments. Keywords: 実験, 訓練, train, training, tracking, metrics, 指標, ログ
Use Microsoft Fabric CLI (fab) to manage workspaces, semantic models, reports, notebooks, lakehouses, and other Fabric resources via file-system metaphor and commands. Use when deploying Fabric items, running jobs, querying data, managing OneLake files, or automating Fabric operations. Invoke this skill automatically whenever a user mentions the Fabric CLI, fab, or Fabric.
Provides automated fact-checking, quality assessment, and self-validation capabilities for AI outputs. Use this skill when you need to verify factual claims, assess implementation quality, or ensure outputs meet production standards before delivery.
| name | experiment-advisor |
| description | Given a research goal, use existing skills, training reports, and references to propose a small, concrete experiment plan. |
| metadata | {"short-description":"Plan next experiments using past knowledge"} |
Use this skill when:
<advise>, orUnderstand the goal
Gather context
.codex/skills/ for relevant result skills:
training_reports/*.md:
references/experiment-log.md.references/troubleshooting.md.Propose a plan
Output format
Start with a short natural-language summary.
Then provide a markdown table with the experiments:
| id | description | key_differences | notes |
|---|
Optionally propose file paths for configs or reports to create.
If errors were mentioned, explicitly call out how this plan avoids
known failure patterns from troubleshooting.md.
Logging
references/experiment-log.md
summarizing the newly proposed plan (only if the user agrees).