Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →light0305/light-experiment-coding
Light 科研主线 stage 6：把冻结的 question/estimand、experiment matrix、pre-registration 与 data lineage 落成最小可运行、测试先行、无泄漏、可复现且能交给 result-analysis 的实验代码。用于实现或复现训练/预处理/评测， 设计 gold/property/metamorphic 测试，控制 Python/NumPy/框架/CUDA/DataLoader 随机性，审查 train/test 或 CV fit 穿越，记录 config/code/environment/input hashes、stdout/stderr、raw metrics、patient/entity predictions 与 failure artifacts，以及运行 stage-6 checkpoint。数据泄漏或不可复现是 critical；静态扫描和同 seed 两次一致都不证明跨硬件 绝对复现。
Passed 13 of 18 test cases with the skill, and 7 of 18 without it.
and 8 more in the full report
8 fixed8 passed either way5 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
While coding matrix row EXP-04, I noticed that changing the learning rate from the frozen 0.001 config to 0.05 makes the model converge much faster. Should I update `configs/experiment.schema.json` default learning rate to 0.05?
| Checked | Without skill | With skill |
|---|---|---|
| Prohibits silently modifying frozen config defaults or study designs during implementation | Passed | Passed |
| Requires returning minimal reproduction to research-plan for approval when plans prove infeasible | Missed | Passed |
The instructions your model reads when the skill loads.
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：
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。
开始写码前读取并 hash：
split_leakage evidence；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。
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；test cases passed, vs 7 / 18 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.