Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/cliffren/swf/papernpx skills add cliffren/swf --skill papergit clone --depth 1 https://github.com/cliffren/swfWrote 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.
[](https://agentmods.dev/skills/cliffren/swf/paper)<a href="https://agentmods.dev/skills/cliffren/swf/paper"><img src="https://agentmods.dev/badge/skills/cliffren/swf/paper.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00014 | $0.01945 |
| Opus 5 | $0.00007 | $0.00972 |
| Sonnet 5 | $0.00003 | $0.00389 |
| Haiku 4.5 | $0.00001 | $0.00194 |
Grade A, and why
paper 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 3d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 203 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Paper Writing Assistant
Generate a paper outline or draft a specific section based on project documentation.
Input
$ARGUMENTS — optional section name. If omitted, generate the full outline.
Valid section names:
| Section | 参数 | 建议写作时机 | 前置依赖 |
|---|---|---|---|
| Methods — 核心设计 | methods-core |
Phase 1 末期(算法确定后) | design.md |
| Methods — 实现细节 | methods-detail |
Phase 2 完成后 | design.md + ADRs + 实验配置 |
| Results — Benchmark | results-benchmark |
Phase 2 完成后 | benchmark 实验记录 |
| Results — Case/Feature | results-case |
Phase 3 完成后 | case study 实验记录 |
| Introduction | intro |
Phase 2~3 间隙 | 文献调研 + outline |
| Discussion | discussion |
Results 写完后 | 全部实验 + design 局限性 |
| Abstract | abstract |
最后写 | 所有其他 section |
| Supplementary | supp |
和主文同步 | 补充图表/方法/数据 |
| Declarations | declarations |
Phase 6 | 作者/基金/数据可用性等 |
| Checklist | checklist |
投稿前 | 检查所有必备项 |
Writing Order and Rationale
推荐顺序(不强制,但每一步都有理由):
1. methods-core ← 最先写,因为算法确定就能写,不依赖实验结果
2. results-benchmark← 跑完 benchmark 趁热写,数据还在脑子里
3. results-case ← 跑完 case study 趁热写
4. methods-detail ← 实验做完后补充数据处理、工具版本、参数选择等细节
5. intro ← 知道自己做出了什么,才能写好 motivation 和 contribution
6. discussion ← 需要看完所有 results 才能讨论
7. abstract ← 全文定型后提炼
8. supp ← 和主文同步,最后整理
Workflow
Generate outline (/swf:paper)
-
Read project docs:
docs/design.md— architecture and core methodsdocs/adr/— key design decisionsdocs/experiments/— results and findingsdocs/experiments/collection-summary.md— if exists, use as primary results sourceCLAUDE.md— project scope and current phase
-
Generate outline and save to
docs/paper/outline.md:# <Project Name>: <Tentative Paper Title> ## Abstract - Key points to cover: ... ## Introduction - Motivation: <why this problem matters> - Gap: <what existing methods can't do> - Contribution: <what we do, 2-3 bullet points> - Paper structure: <brief roadmap of sections> ## Methods ### Core Design - <Subsection 1 from design.md>: key concepts, core formulas - <Subsection 2>: ... ### Implementation Details - Data preprocessing: ... - Training/optimization: ... - Tools and versions: ... ## Results ### Benchmark - Datasets: <which ones> - Baselines: <which methods> - Metrics: <which metrics> - Key findings: <1-2 sentences> - Figures: Fig.X — <description> ### Case Study / Feature Validation - Case 1: <dataset, question, finding> - Case 2: ... - Figures: Fig.X — <description> ### Performance - Runtime, memory, scalability - Figure: Fig.X — <description> ## Discussion - Interpretation of key results - Comparison with related work - Limitations - Future directions ## Supplementary - Supplementary figures - Supplementary methods - Supplementary tables ## Declarations (投稿前必备) - Author contributions: <每位作者的具体贡献> - Competing interests: <每位作者的利益冲突声明> - Funding: <基金名称 + 编号> - Data availability: <数据存放位置、获取方式> - Code availability: <repo URL + license + version/commit> - Ethics approval: <如涉及> - Acknowledgments: <致谢> ## Metadata - Authors & affiliations - Corresponding author (email, ORCID) - Keywords (3-7) - Cover letter outline
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.
- 3d ago First seen · 203 lines · 14 tokens per session scan A 5c82bcfd1e7b
paper is a skill published in the GitHub repository cliffren/swf (5 stars, last pushed 4mo ago), licensed MIT. It adds 14 tokens to every session and 1,945 once invoked, about $0.0001 per session on Opus 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 2026-08-31.
Other skills, from other repositories
exploratory-data-analysis
Perform bounded, local exploratory analysis of explicitly supported scientific files. Use for redacted CSV/TSV/JSON profiles; optional NumPy, HDF5, FASTA/FASTQ, and basic image metadata inspection; missingness/leakage audits; outlier and transformation sensitivity; and rigorous EDA report scaffolds. Other domain…
nature-statistics
Audit, revise, or draft manuscript statistical reporting for Nature / high-impact journal submissions. Use when the user asks to check statistical analysis sections, p values, confidence intervals, sample size, biological versus technical replicates, randomization, blinding, multiple-comparison correction, model…
evaluating-with-leakage-gates
Evaluate an OpenMed de-identification or clinical NER model against the leakage-first release gates G1a through G8, which gate releases on residual PHI leakage rather than on F1. Use when the user wants to run the OpenMed eval harness on a synthetic golden set, decide whether a de-id model is RELEASABLE or…
mapping-to-snomed
Maps clinical concept spans extracted by OpenMed to SNOMED CT concepts through a USER-SUPPLIED terminology server (the user's own Ontoserver, Snowstorm, or UMLS/UTS), never a bundled vocabulary. Use when the user wants to code findings, disorders, procedures, body structures, or substances to SNOMED CT, run an ECL…
mixed-precision
Use FP16/BF16 mixed precision to accelerate training and reduce memory. Use when optimizing GPU performance.
indication-dossier
Build a source-backed biomedical indication dossier. Use when a research task asks for disease biology, target rationale, patient segmentation, biomarkers, trials, drugs, competitive landscape, or translational evidence.