joshua-zyy/academic-paper-writer / skills/academic-experiments
academic-experiments
VerifiedAudit, run, or verify experimental evidence for CS/AI/ML papers. Produces Evidence Inventory with evidence_type annotations (newly_run/preexisting_artifact/user_claim) and Protocol Risk assessments. Use when: checking if experiment results are reproducible, auditing existing experiment artifacts, running minimal reproducible commands, evaluating checkpoints without full retraining, documenting protocol risks like data leakage or missing baselines. Triggers on: 复核实验, run experiments, 实验结果, experiment evidence, verify results, 实验验证, evidence inventory, protocol risk, 跑实验, check results, reproduce experiments, 实验审计.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.1 KB |
| agents/experiment_agent.md | markdown | 5.9 KB |
| manifest.yaml | config | 1.2 KB |
| references/evidence-inventory.md | markdown | 1.7 KB |
| references/protocol-risks.md | markdown | 1.8 KB |
| references/run-strategy.md | markdown | 1.4 KB |
| scripts/evidence_scanner.py | script | 4.2 KB |
| static/core/anti-patterns.md | markdown | 571 B |
| static/core/output-contract.md | markdown | 906 B |
| static/core/red-lines.md | markdown | 477 B |
| static/core/stance.md | markdown | 2.5 KB |
references/run-strategy.md
View on GitHub# 运行策略与记录格式
## 运行优先级
默认按以下顺序执行,而非"一上来就 full training":
1. **验证环境** — 检查 Python/CUDA 版本、依赖是否安装
2. **最小可复核命令** — 评估已有 checkpoint 或解释脚本
3. **评估或解释** — 运行 eval.py 或解释/可视化脚本
4. **重训** — 只有在确有必要时才进行完整训练
## 优先动作
- 评估现有 `best_model.pth`
- 运行解释脚本验证图表来源
- 复核单个 split 的指标计算流程
- 检查 subject-level 与 file-level 划分差异
## 运行记录格式
对每个进入论文正文的实验结果,至少记录:
```md
## Experiment Evidence
- Status: newly_run / preexisting_artifact / blocked
- Command: (实际执行的命令)
- Workdir: (工作目录)
- Environment: (Python 版本、CUDA 版本、关键依赖)
- Inputs: (输入数据、checkpoint 路径)
- Key Config: (关键超参数)
- Output Artifacts: (输出文件路径)
- Metrics Used In Draft: (正文中引用的指标值)
- Protocol Risks: (见 protocol-risks.md)
```
## 常见错误
以下行为会破坏证据链:
- 没区分重跑结果和仓库自带结果
- 没核对指标定义就把数字抄进论文
- 把 file-level split 写成 subject-level split
- 把验证集调阈值后的结果写成最终测试性能
- 只看到图,却不知道图对应哪个脚本和哪个 checkpoint
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |