agentmods.dev

light-experiment-coding

light-experiment-coding is a skill for Claude Code, Codex from Light0305/Light-skills. It costs 170 tokens per session (3,305 once invoked), scanned A, original, MIT.

A workflow for turning a fixed research experiment plan into small, test-first, reproducible code. It covers training, preprocessing, evaluation, data lineage, randomness, and protection against information leaking from test data into training.

In plain words
What is it for?
Use it to implement or reproduce experiments, create unit and property-based tests, control seeds and environments, detect train/test leakage, and preserve inputs, outputs, logs, and hashes for later analysis.
Why use it?
It helps ensure that experiments answer the planned question fairly and can be rerun and audited. TDD means test-driven development: writing checks before or alongside the implementation.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Needs its repository: it reads a path above its own folder, which exists only inside the repository. The line is python ../light-orchestrator/scripts/run_checkpoint.py \.

not rated 683repo +42 3mo ago A scan SkillSpector: pass 170 tokens original MIT

Good fit Use it to implement or reproduce experiments, create unit and property-based tests, control seeds and environments, detect train/test leakage, and preserve inputs, outputs, logs, and hashes for later analysis.

Compare 6 skills from other repositories ↓
View source ↗ Light0305/Light-skills
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/Light0305/Light-skills
agentmods
npx agentmods add skills/light0305/light-skills/light-experiment-coding

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for light-experiment-coding

README.md
[![agentmods](https://agentmods.dev/badge/skills/light0305/light-skills/light-experiment-coding/github.svg)](https://agentmods.dev/skills/light0305/light-skills/light-experiment-coding)
Your own site
<a href="https://agentmods.dev/skills/light0305/light-skills/light-experiment-coding"><img src="https://agentmods.dev/badge/skills/light0305/light-skills/light-experiment-coding/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for light-experiment-coding

Your own site · 80×15
<a href="https://agentmods.dev/skills/light0305/light-skills/light-experiment-coding"><img src="https://agentmods.dev/badge/skills/light0305/light-skills/light-experiment-coding.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 170 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,305 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00170 $0.03305
Opus 5.5 $0.00068 $0.01322
Sonnet 5.5 $0.00034 $0.00661
Haiku 4.5 $0.00017 $0.00331

Measured 1mo ago against content hash 2cdad9b18f3c, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

light-experiment-coding scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.

The scan reads SKILL.md. This mod also ships 15 executable files (assets/project-scaffold/scripts/boundary_trace.py, assets/project-scaffold/scripts/debug_instrument.sh, assets/project-scaffold/src/example/__init__.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/light-experiment-coding/SKILL.md · 208 lines

How it starts

The opening of the file, as written. The whole thing — 208 lines — stays where its author put it; the contents beside it link to each section on GitHub.

实验编码（stage 6）

任务不是“写出能跑的 notebook”，而是把上游冻结计划逐行实现成可证伪、可复跑、可审计的实验。优先级：

  1. 不改研究问题；
  2. 不让评估信息进入训练；
  3. 能用固定环境与 seed 真复跑；
  4. 保留足够 raw evidence，让 result-analysis 自己重算；
  5. 代码整洁和速度服从以上约束。

先完整阅读 references/experiment-coding-resource-map.md。工具机制见 references/tools.md，TDD/调试红旗见 references/tdd_redflags.md 与 references/debug_protocol.md。

入口：冻结输入

开始写码前读取并 hash：

  • question / estimand；
  • experiment matrix 每一行和 fair-comparison 常量；
  • pre-registration 及 provenance；
  • failure-tree report：每条 hypothesis 的 success/failure/inconclusive 分支、guardrail/counter-metric、kill criterion 与 amendment policy；
  • data fixed revision、raw/curated SHA256、lineage、split ID、split_leakage evidence；
  • result-analysis 对 raw run、predictions、metrics、failures、provenance 的消费契约；
  • 当前 git commit 与 dirty state。

primary outcome、comparison family、exclusion、stopping 已冻结。若实现证明计划不可行，带最小复现和影响返回 research-plan，停下让人决策；不得改 config 默认值静默漂移。

实现顺序

1. 建立最小可运行项目

优先复制 assets/project-scaffold/：

  • uv.lock + pyproject.toml：uv sync --locked --extra dev；
  • configs/experiment.schema.json：每个 matrix row 的机读配置；
  • experiment_contracts.py：data/model/metric/preprocessing 最小接口；
  • reproducibility.py：运行期 seed helper；
  • CI/pre-commit/debug 资产。

遵循现有仓库框架和配置格式；不要为一个实验引入付费 IDE、云追踪或私有 key。MLflow/DVC 可选，普通本地文件必须能完成 核心闭环。不可用资源明确写 UNAVAILABLE:原因，不假装通过。

2. 测试先行

在实现 preprocessing/train/eval 前：

  1. 写 gold test，验证人工可算的小答案；
  2. 写 property test（Hypothesis），验证范围、有限性、对称/单调等不变量；
  3. 写 metamorphic test，验证置换/等价变换后的输出关系；
  4. 写 train-only-fit 测试，记录 transformer 只收到训练折；
  5. 亲眼看新测试因缺实现或真实 bug 失败，再写最小实现使其通过。

浮点断言用 pytest.approx / assert_allclose(rtol, atol)。先声明 device、dtype、mixed precision 和容差；NaN/Inf 默认 fail。不要测“随机训练一定达到某个漂亮数”，测确定性边界和可重算事实。

3. 防泄漏实现

  • holdout：先 split，再仅用 train fit/fit_transform，test 只 transform；
  • CV/调参：预处理器与模型放进 sklearn Pipeline，每折只 fit training fold；
  • 患者/用户/牧场等实体用 group-aware split，不得跨 train/test；
  • 目标编码、特征选择、PCA、imputation 同样只在训练折 fit；
  • 数据件复核直接复用 data-engineering split_leakage，不重造。

Read the full file on GitHub · 208 lines

Files

What ships with it

32 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 1mo ago First seen · 208 lines · 170 tokens per session scan A 2cdad9b18f3c

Subscribe to this mod's changes

light-experiment-coding is a skill published in the GitHub repository Light0305/Light-skills (683 stars, last pushed 3mo ago), licensed MIT. It adds 170 tokens to every session and 3,305 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Claude CodePaper WritingScientificTestingacademic-writingai-agentcodexllm-skills
Related

Other skills, from other repositories

anti-defensive-writing-en

Stops defensive writing across the entire paper lifecycle — writing, revising, cutting, and organizing experiments. Treats the paper as a press conference, not a project summary, lab log, or self-audit: identify the single most publishable strength of the work and build the most favorable, complete, and persuasive…

Adkid-Zephyr/anti-defensive-writing-Skill · 138 tokens

simulation-study

Scaffold and run a reproducible Monte Carlo simulation study in R — a declared assumption regime, a parameterized DGP, an estimator grid, a seeded replication loop, and a summary of bias, RMSE, empirical SE, coverage, size/power with Monte Carlo standard errors. Use when the user says "run a Monte Carlo simulation"…

pedrohcgs/claude-code-my-workflow · 145 tokens

anti-defensive-writing

A Chinese-language writing guide for presenting a research paper around its strongest supported contribution. It treats the paper as a focused academic presentation rather than a project diary or complete lab record.

Adkid-Zephyr/anti-defensive-writing-Skill · 160 tokens

academic-citation

Search, verify, and map citations for CS/AI/ML papers. Produces VERIFIED/UNVERIFIED reference lists with Citation-to-Claim maps and Exemplar Sets. Use when: finding references for a paper section, verifying citation accuracy, building exemplar sets for introduction/related work learning, checking if existing citations…

joshua-zyy/academic-paper-writer · 146 tokens

academic-experiments

Audit, run, or verify experimental evidence for CS/AI/ML papers. Produces Evidence Inventory with evidencetype annotations (newlyrun/preexistingartifact/userclaim) and Protocol Risk assessments. Use when: checking if experiment results are reproducible, auditing existing experiment artifacts, running minimal…

joshua-zyy/academic-paper-writer · 132 tokens

academic-figure

Create, revise, or audit academic data/result figures for CS/AI/ML papers. Data/result plots default to Python-generated editable SVG with CS/AI/ML-specific design rules for benchmarks, ablations, training dynamics, robustness, diagnostics, distributions, confusion matrices, and efficiency tradeoffs. Use when…

joshua-zyy/academic-paper-writer · 235 tokens