dsh-aris-panel 0.1.0

package/skills/writing-systems-papers/SKILL.md

@@ -0,0 +1,184 @@
+---
+name: writing-systems-papers
+description: "Paragraph-level structural blueprint for 10-12 page systems papers targeting OSDI, SOSP, ASPLOS, NSDI, and EuroSys. Provides page allocation, paragraph templates, and writing patterns. Use when user says \"写系统论文\", \"systems paper structure\", \"OSDI paper\", \"SOSP paper\", or wants fine-grained structural guidance for a systems conference submission."
+argument-hint: "[venue-or-section]"
+allowed-tools: Bash(*), Read, Write, Edit, Grep, Glob, WebSearch, WebFetch, mcp__codex__codex, mcp__codex__codex-reply
+---
+
+# Writing Systems Papers: Paragraph-Level Blueprint
+
+Structural guidance for **$ARGUMENTS**
+
+## Relationship to Other ARIS Skills
+
+- **paper-write**: General paper generation workflow with citation verification. This skill complements it with systems-specific structural blueprints.
+- **paper-slides**: Conference presentation generation (Beamer+PPTX). Already covers talks — no overlap.
+- **paper-plan**: Research outline creation. Use paper-plan first, then this skill for paragraph-level structure.
+
+**Boundary**: paper-write handles the generation workflow (LaTeX output, DBLP verification, section-by-section drafting). This skill provides the **structural skeleton** — page budgets, paragraph roles, and writing patterns specific to systems venues.
+
+---
+
+## Page Allocation: 12-Page Systems Paper
+
+| Section | Pages | Key Content |
+|---------|-------|-------------|
+| Abstract | ~0.25 | 150–250 words, 5 sentences |
+| S1 Introduction | 1.5–2 | Problem → Gap → Insight → Contributions |
+| S2 Background & Motivation | 1–1.5 | Terms + Production observations |
+| S3 Design | 3–4 | Architecture + Modules + Alternatives |
+| S4 Implementation | 0.5–1 | Prototype, LOC, engineering |
+| S5 Evaluation | 3–4 | Setup + E2E + Ablation + Scalability |
+| S6 Related Work | 1 | By methodology, explicit comparison |
+| S7 Conclusion | 0.5 | 3-sentence summary |
+
+---
+
+## Section Blueprints
+
+### Abstract (5 sentences)
+
+```text
+S1: Problem context and importance
+S2: Gap in existing approaches
+S3: Thesis — "X is better for Y in environment Z" (Irene Zhang formula)
+S4: Approach summary + headline results
+S5: Impact or availability
+```
+
+Sources: Levin & Redell — "Can you state the new idea concisely?"; Irene Zhang — "abstract cannot use terms introduced in the paper."
+
+### S1 Introduction (1.5–2 pages)
+
+1. **Problem** (~0.5p) — Domain + concrete numbers + why it matters
+2. **Gap analysis** (~0.5p) — G1–Gn: specific shortcomings with evidence
+3. **Key insight** (1 para) — Thesis: "X is better for Y in Z"
+4. **Contributions** (~0.5p) — 3–5 numbered, testable claims with §N references
+
+Pattern: hzwer Move 1 (territory) → Move 2 (niche) → Move 3 (occupy).
+
+### S2 Background & Motivation (1–1.5 pages)
+
+1. **Technical background** (~0.5p) — Define-before-use (Gernot Heiser)
+2. **Observations** (~0.5–1p) — O1, O2, O3 from production data → design insights
+
+### S3 Design (3–4 pages)
+
+1. **Architecture overview** (~0.5p) — Diagram first (Yi Ding: "draw a picture first")
+2. **Module details** (~2–2.5p) — Per module: choice, alternatives, why
+3. **Trade-offs** (~0.5–1p) — Summary of design decisions
+
+Rule: "Every design choice must discuss alternatives" (Irene Zhang).
+
+### S4 Implementation (0.5–1 page)
+
+Language, LOC, framework, key engineering decisions. Keep concise.
+
+### S5 Evaluation (3–4 pages)
+
+1. **Setup** (~0.5p) — Hardware, baselines, workloads, metrics
+2. **End-to-end** (~1–1.5p) — X vs baselines for Y on Z
+3. **Ablation** (~1–1.5p) — Remove each component, measure impact
+4. **Scalability** (~0.5p) — Behavior at increasing scale
+
+**Three-statement rule** (Irene Zhang): Every conclusion stated as:
+- Hypothesis (section opening)
+- Conclusion (section closing)
+- Caption (figure caption)
+
+### S6 Related Work (1 page)
+
+Group by methodology. For each group: what they do, limitation, how we differ.
+
+### S7 Conclusion (0.5 page)
+
+Three sentences: problem, solution, result. No new information.
+
+---
+
+## Writing Patterns
+
+### Pattern 1: Gap Analysis
+Enumerate G1–Gn in intro → A1–An in design → verify in evaluation.
+*Example*: Lucid (ASPLOS'23) — 5 gaps mapped to 5 answers.
+
+### Pattern 2: Observation-Driven
+O1–O3 from production data → insights → design components.
+*Example*: GFS (arXiv 2025) — 3 observations drive 3 components.
+
+### Pattern 3: Contribution List
+Numbered contributions in intro, each with §N cross-reference.
+*Example*: Blox (EuroSys'24) — 7 contributions; Sia (SOSP'23) — 5 contributions.
+
+### Pattern 4: Thesis Formula
+"X is better for Y in Z" structures the entire paper.
+Combine with other patterns for maximum impact.
+
+---
+
+## Conference Differences
+
+> Always verify against current CFP — rules change yearly.
+
+| Venue | Format | Pages | Camera-Ready |
+|-------|--------|-------|-------------|
+| OSDI | USENIX | 12 | 14 |
+| NSDI | USENIX | 12 | 14 |
+| SOSP | ACM SIGOPS | 12 | — |
+| ASPLOS | ACM SIGPLAN | 11 | 13 |
+| EuroSys | ACM | 12 | — |
+
+Based on 2025/2026 CFPs.
+
+---
+
+## Workflow
+
+```text
+1. Determine venue and page limit
+2. Choose writing pattern (Gap/Observation/Contribution/Thesis)
+3. Allocate pages per section using the table above
+4. Draft Abstract following 5-sentence template
+5. Draft Introduction: Problem → Gap → Insight → Contributions
+6. Draft Motivation with production observations (if available)
+7. Draw architecture figure, then write Design
+8. Draft Implementation (concise)
+9. Draft Evaluation: setup → E2E → ablation → scalability
+10. Draft Related Work by methodology groups
+11. Draft Conclusion: 3 sentences
+12. Run pre-submission checklist
+13. Hand off to /paper-write for LaTeX generation and citation verification
+```
+
+---
+
+## Quick Self-Check
+
+- [ ] Thesis follows "X is better for Y in Z"
+- [ ] 3–5 numbered contributions with §N references
+- [ ] Design discusses alternatives for every major choice
+- [ ] Eval conclusions stated 3 times (hypothesis, result, caption)
+- [ ] Related work grouped by methodology
+- [ ] Page budget within venue limits
+- [ ] No fabricated observations, traces, or results
+- [ ] All citations verified (delegate to /paper-write)
+
+---
+
+## Academic Integrity
+
+- Never fabricate observations, traces, or experimental results
+- Never generate citations from memory — use /paper-write citation workflow
+- Disclose LLM use per venue policy
+- This blueprint provides structural guidance, not copy-paste text
+
+---
+
+## Authoritative Sources
+
+1. Levin & Redell — "How (and How Not) to Write a Good Systems Paper" (USENIX)
+2. Irene Zhang — "Hints on how to write an SOSP paper"
+3. Gernot Heiser — Style Guide + Paper Writing Talk
+4. Timothy Roscoe — "Writing reviews for systems conferences"
+5. Yi Ding — "How to write good systems papers?"
+6. hzwer & DingXiaoH — WritingAIPaper (GitHub)

package/templates/CLAUDE_MD_TEMPLATE.md

@@ -0,0 +1,29 @@
+# Project: {project-name}
+
+## Pipeline Status
+
+```yaml
+stage: idle          # idle | idea-discovery | implementation | training | review | paper
+idea: ""             # Current idea title (one line)
+contract: ""         # Path to research_contract.md (e.g., idea-stage/docs/research_contract.md)
+current_branch: ""   # Git branch for this idea
+baseline: ""         # Baseline numbers for comparison
+training_status: idle  # idle | running | complete | failed
+language: en         # en | zh — controls skill output language (see shared-references/output-language.md)
+active_tasks: []
+next: ""             # Concrete next step
+last_updated: ""     # YYYY-MM-DD HH:mm — auto-updated by skills on every output write
+```
+
+## Project Constraints
+
+- {constraint 1}
+- {constraint 2}
+
+## Non-Goals
+
+- {non-goal 1}
+
+## Compute Budget
+
+- {budget details, e.g., "8x A100 for 24h via vast.ai"}

package/templates/EXPERIMENT_LOG_TEMPLATE.md

@@ -0,0 +1,47 @@
+# Experiment Log
+
+> **Complete record of all experiments run in this project.** Every experiment gets an entry — successful or not. This is the authoritative source for "what did we actually run and what happened?"
+>
+> **How it differs from EXPERIMENT_TRACKER.md:** The tracker (in `refine-logs/`) is an execution checklist (TODO → RUNNING → DONE). This log is a permanent record with full results, configs, and reproduction commands. The tracker tells you what's left to do; the log tells you what was done and what it showed.
+>
+> **Update rule:** Write an entry immediately after each experiment completes. Do not batch entries or wait until "later."
+
+## Experiment: [Descriptive Name]
+
+**Date**: YYYY-MM-DD
+**Idea**: [Which idea from IDEA_CANDIDATES.md]
+**Goal**: [What this experiment tests — link to claim if applicable]
+
+### Setup
+- **Method**: [Brief description of the approach]
+- **Dataset**: [Name, split, size]
+- **Baseline**: [What you compare against]
+- **Hardware**: [Server, GPUs, time taken]
+- **Config**: [Path to config file or key hyperparameters]
+
+### Results
+
+| Method | Dataset | Metric-1 | Metric-2 | Notes |
+|--------|---------|----------|----------|-------|
+| Baseline | [dataset] | [number] | [number] | [reproduced / from paper] |
+| Ours | [dataset] | [number] | [number] | [seeds, std if applicable] |
+
+### Verdict
+- **Supports claim?** [Yes / Partially / No]
+- **Key takeaway**: [One sentence — what did we learn?]
+
+### Reproduction
+```bash
+# Command to reproduce this experiment
+python train.py --config configs/exp01.yaml --seed 42
+```
+
+### WandB
+- Run URL: [link]
+- Run ID: [id]
+
+---
+
+## Experiment: [Next Experiment Name]
+
+...

package/templates/EXPERIMENT_PLAN_TEMPLATE.md

@@ -0,0 +1,51 @@
+# Experiment Plan
+
+> **Template for Workflow 1.5 (`/experiment-bridge`).** Fill in, save as `refine-logs/EXPERIMENT_PLAN.md`, then run `/experiment-bridge`.
+
+**Problem**: [What problem does your method solve?]
+**Method Thesis**: [One-sentence description of your approach]
+
+## Claim Map
+
+| Claim | Why It Matters | Minimum Convincing Evidence | Linked Blocks |
+|-------|----------------|----------------------------|---------------|
+| C1: [Main claim] | [Why] | [Evidence needed] | B1, B2 |
+| C2: [Supporting claim] | [Why] | [Evidence needed] | B3 |
+
+## Experiment Blocks
+
+### Block 1: Main Result
+- **Claim tested**: C1
+- **Dataset / split / task**: [e.g., ImageNet val]
+- **Compared systems**: [Your method vs. Baseline A vs. Baseline B]
+- **Metrics**: [Primary: accuracy/PPL. Secondary: throughput]
+- **Setup details**: [Backbone, optimizer, lr, epochs, seeds]
+- **Success criterion**: [e.g., "> 2% accuracy over baseline"]
+- **Failure interpretation**: [If negative, what does it mean?]
+- **Priority**: MUST-RUN
+
+### Block 2: Ablation Study
+- **Claim tested**: C1 (novelty isolation)
+- **Compared systems**: [Full method, -component A, -component B]
+- **Success criterion**: [Each component contributes > 0.5%]
+- **Priority**: MUST-RUN
+
+### Block 3: [Additional Experiment]
+- **Priority**: NICE-TO-HAVE
+
+## Run Order
+
+| Milestone | Goal | Runs | Decision Gate | Cost |
+|-----------|------|------|---------------|------|
+| M0: Sanity | Pipeline works | 1 quick run | Loss decreases? | ~0.5h |
+| M1: Baselines | Reproduce baselines | Block 3 | Numbers match? | ~4h |
+| M2: Main | Full method | Block 1 | Meets criterion? | ~8h |
+| M3: Ablation | Components | Block 2 | Each matters? | ~6h |
+
+## Compute Budget
+- **Total estimated GPU-hours**: ~18h
+- **Hardware**: [e.g., 4x RTX 3090]
+- **Biggest bottleneck**: [e.g., baseline reproduction]
+
+## Risks
+- **Risk**: [What could go wrong] → **Mitigation**: [How to handle it]

package/templates/EXPERIMENT_PLAN_TEMPLATE_CN.md

@@ -0,0 +1,53 @@
+# 实验计划
+
+> **Workflow 1.5 (`/experiment-bridge`) 的模板。** 填写后保存为 `refine-logs/EXPERIMENT_PLAN.md`，然后运行 `/experiment-bridge`。
+
+**问题**: [你的方法解决什么问题？]
+**方法论点**: [一句话描述你的方法]
+
+## Claim 映射
+
+| Claim | 重要性 | 最低说服力证据 | 关联实验块 |
+|-------|--------|---------------|-----------|
+| C1: [主要 claim] | [为什么重要] | [需要的证据] | B1, B2 |
+| C2: [辅助 claim] | [为什么重要] | [需要的证据] | B3 |
+
+## 实验块
+
+### 实验块 1: 主实验
+- **验证 Claim**: C1
+- **数据集 / 划分 / 任务**: [如：ImageNet val]
+- **对比系统**: [你的方法 vs. 基线 A vs. 基线 B]
+- **评估指标**: [主要：准确率/PPL。次要：吞吐量]
+- **实验设置**: [骨干网络、优化器、学习率、训练轮数、随机种子]
+- **成功标准**: [如："> 基线 2% 准确率"]
+- **失败解读**: [如果结果为负，意味着什么？]
+- **优先级**: 必须运行
+
+### 实验块 2: 消融实验
+- **验证 Claim**: C1（新颖性隔离）
+- **对比系统**: [完整方法、去掉组件 A、去掉组件 B]
+- **成功标准**: [每个组件贡献 > 0.5%]
+- **优先级**: 必须运行
+
+### 实验块 3: [额外实验]
+- **优先级**: 可选
+
+## 运行顺序
+
+| 里程碑 | 目标 | 运行内容 | 决策关卡 | 预估耗时 |
+|--------|------|---------|---------|---------|
+| M0: 健全性检查 | 流程跑通 | 1 次快速运行 | loss 在下降？ | ~0.5h |
+| M1: 基线复现 | 复现基线结果 | 实验块 3 | 数值匹配？ | ~4h |
+| M2: 主实验 | 完整方法 | 实验块 1 | 达到标准？ | ~8h |
+| M3: 消融 | 组件验证 | 实验块 2 | 每个都有贡献？ | ~6h |
+
+## 算力预算
+
+- **预估总 GPU 小时**: ~18h
+- **硬件**: [如：4x RTX 3090]
+- **最大瓶颈**: [如：基线复现]
+
+## 风险
+
+- **风险**: [可能出什么问题] → **缓解措施**: [如何应对]

package/templates/FINDINGS_TEMPLATE.md

@@ -0,0 +1,52 @@
+# Findings
+
+> **Cross-stage discovery log.** Records what you learn during experiments — both research insights about your method/claims and engineering lessons from debugging. Read on every session recovery, so keep entries concise.
+>
+> **Why this file exists:** Experiments produce discoveries that are critical for future decisions but don't belong in formal experiment reports. Without a central log, these get lost between sessions — and the next session repeats the same mistakes or misses important signals.
+
+---
+
+# Research Findings
+
+> Method-level insights: what works, what doesn't, and why. These directly inform your claims, experiment design, and paper narrative.
+
+## [YYYY-MM-DD] Topic
+- Finding
+- Evidence (wandb run, metric, dataset)
+
+## [YYYY-MM-DD] Example: Attention module ineffective on small datasets
+- Our proposed attention mechanism shows no improvement on CIFAR-10 (acc 93.1 vs baseline 93.0) but +2.3% on ImageNet
+- Hypothesis: the module needs sufficient spatial diversity to capture meaningful patterns; small-resolution inputs don't provide this
+- Implication: restrict claims to medium/large-scale datasets
+
+## [YYYY-MM-DD] Example: Loss combination causes gradient explosion
+- Combining L_contrastive + L_distill with equal weight → gradient norm >1000 after epoch 5
+- Root cause: L_contrastive scale is ~10x larger than L_distill; needs rebalancing or gradient clipping
+- Decision: weight ratio 0.1:1.0, gradient norm stable at ~5.0 after fix
+
+## [YYYY-MM-DD] Example: Key decision — dropped multi-scale approach
+- Multi-scale variant adds 40% compute but only +0.3% accuracy over single-scale
+- Not worth the complexity; claim reframed around efficiency rather than raw performance
+
+---
+
+# Engineering Findings
+
+> Infrastructure, environment, and debugging lessons. Prevents re-debugging the same issues in future sessions.
+
+## [YYYY-MM-DD] Topic
+- Problem and root cause
+- Fix applied
+
+## [YYYY-MM-DD] Example: OOM with gradient accumulation
+- OOM on batch_size=32 with 4x GPU — gradient accumulation was doubling peak memory
+- Fix: enabled gradient checkpointing; batch_size=32 now fits in 24GB
+
+## [YYYY-MM-DD] Example: Baseline reproduction gap
+- Paper reports 95.5 on dataset-X; we get 93.2 with their official code
+- Root cause: different data preprocessing (center crop vs resize-then-crop)
+- Decision: use our preprocessing for fair comparison, document in paper
+
+## [YYYY-MM-DD] Example: WandB logging breaks DDP
+- wandb.log() inside DDP forward pass causes hanging on multi-GPU
+- Fix: wrap in `if dist.get_rank() == 0` guard

package/templates/IDEA_CANDIDATES_TEMPLATE.md

@@ -0,0 +1,47 @@
+# Idea Candidates
+
+> **Curated pool of viable ideas that survived novelty check and review.** Unlike `IDEA_REPORT.md` (raw brainstorm output with all 8-12 ideas), this file only contains ideas that are worth pursuing — filtered, scored, and ready to be selected.
+>
+> When the current idea fails (result-to-claim negative, or pivot needed), pick the next candidate from here instead of re-running the entire idea discovery pipeline.
+
+## Active Idea
+
+**Current**: [Idea name] → see `docs/research_contract.md` for full context
+
+## Candidates
+
+### Candidate 1: [Idea Name]
+- **One-line**: [What it does]
+- **Novelty score**: X/10
+- **Reviewer assessment**: [Key strength + key risk]
+- **Pilot result**: [POSITIVE/NEGATIVE/NOT TESTED — one-line metric]
+- **Estimated effort**: [GPU hours, implementation complexity]
+- **Why not selected first**: [Reason — e.g., higher risk, needs more data, less novel]
+
+### Candidate 2: [Idea Name]
+- **One-line**: [What it does]
+- **Novelty score**: X/10
+- **Reviewer assessment**: [Key strength + key risk]
+- **Pilot result**: [POSITIVE/NEGATIVE/NOT TESTED]
+- **Estimated effort**: [GPU hours, implementation complexity]
+- **Why not selected first**: [Reason]
+
+### Candidate 3: [Idea Name]
+...
+
+## Killed Ideas
+
+> Ideas that were evaluated and explicitly rejected. Kept here to avoid re-exploring dead ends.
+
+### [Killed Idea Name]
+- **Kill reason**: [Why — e.g., not novel (found prior work X), infeasible (needs Y data), pilot negative]
+- **Kill date**: YYYY-MM-DD
+- **Source**: IDEA_REPORT.md, Idea #N
+
+## Idea Switch Log
+
+> Track idea transitions to maintain context across switches.
+
+| Date | From | To | Reason |
+|------|------|----|--------|
+| YYYY-MM-DD | [Previous idea] | [New idea] | [Why — e.g., result-to-claim negative, pivot after reviewer feedback] |

package/templates/IDEA_CANDIDATES_TEMPLATE_CN.md

@@ -0,0 +1,47 @@
+# Idea 候选池
+
+> **经过新颖性检查和评审后存活的可行 idea 精选池。** 与 `idea-stage/IDEA_REPORT.md`（包含全部 8-12 个 idea 的原始头脑风暴输出）不同，本文件只包含值得推进的 idea — 已筛选、已评分、可直接选用。
+>
+> 当前 idea 失败时（result-to-claim 为负，或需要转向），从这里选择下一个候选，而不是重新运行整个 idea 发现流程。
+
+## 当前 Idea
+
+**进行中**: [Idea 名称] → 详见 `idea-stage/docs/research_contract.md`
+
+## 候选列表
+
+### 候选 1: [Idea 名称]
+- **一句话描述**: [做什么]
+- **新颖性评分**: X/10
+- **评审意见**: [核心优势 + 核心风险]
+- **试点结果**: [正面/负面/未测试 — 一行指标]
+- **预估工作量**: [GPU 小时数，实现复杂度]
+- **未优先选择的原因**: [原因 — 如：风险较高、需要更多数据、新颖性较低]
+
+### 候选 2: [Idea 名称]
+- **一句话描述**: [做什么]
+- **新颖性评分**: X/10
+- **评审意见**: [核心优势 + 核心风险]
+- **试点结果**: [正面/负面/未测试]
+- **预估工作量**: [GPU 小时数，实现复杂度]
+- **未优先选择的原因**: [原因]
+
+### 候选 3: [Idea 名称]
+...
+
+## 已淘汰 Idea
+
+> 经过评估后明确拒绝的 idea。保留在此以避免重复探索死胡同。
+
+### [已淘汰 Idea 名称]
+- **淘汰原因**: [为什么 — 如：不够新颖（发现先前工作 X）、不可行（需要 Y 数据）、试点结果为负]
+- **淘汰日期**: YYYY-MM-DD
+- **来源**: idea-stage/IDEA_REPORT.md, Idea #N
+
+## Idea 切换日志
+
+> 跟踪 idea 转换，在切换过程中保持上下文。
+
+| 日期 | 从 | 到 | 原因 |
+|------|----|----|------|
+| YYYY-MM-DD | [之前的 idea] | [新 idea] | [为什么 — 如：result-to-claim 为负、评审反馈后转向] |

package/templates/INVENTION_BRIEF_TEMPLATE.md

@@ -0,0 +1,87 @@
+# Invention Brief
+
+> **Template for input to `/patent-pipeline`.** Provide detailed invention disclosure instead of a one-line prompt. Supports both structured and conversational input.
+
+## Title (发明名称)
+[A concise title reflecting the technical content. e.g., "一种基于注意力机制的图像去噪方法及装置"]
+
+## Technical Field (技术领域)
+[1-2 paragraphs: What technical domain does this invention belong to?]
+
+## Inventors
+- [Name 1 — role/contribution]
+- [Name 2 — role/contribution]
+
+## Background — Known Prior Art (背景技术)
+- **Closest prior art**: [Known patents, papers, or products]
+- **What they do well**: [Strengths of existing approaches]
+- **Their deficiencies**: [Specific technical problems not solved]
+- **Gap**: [What is missing that this invention addresses]
+
+## Technical Problem (要解决的技术问题)
+[State the specific technical problem this invention solves, derived from prior art deficiencies.]
+
+## Proposed Solution (技术方案)
+[Describe the invention's technical solution at a high level. What are the core inventive features?]
+
+## Key Advantages (有益效果)
+[Quantifiable or measurable advantages over prior art.]
+- Advantage 1: [description + evidence if available]
+- Advantage 2: [description + evidence if available]
+- Advantage 3: [description + evidence if available]
+
+## Embodiments (实施例)
+[Describe at least one way to practice the invention. Include specific parameters, components, steps, or configurations.]
+
+### Embodiment 1: [Title]
+[Detailed description]
+
+### Embodiment 2: [Title — optional]
+[Detailed description]
+
+## Figures Available (附图)
+[List available figures. For each figure, briefly describe what it shows.]
+
+| Figure | Description | File |
+|--------|-------------|------|
+| Fig. 1 | [e.g., system architecture block diagram] | [path/to/fig1.png] |
+| Fig. 2 | [e.g., method flowchart] | [path/to/fig2.png] |
+| Fig. 3 | [e.g., experimental comparison chart] | [path/to/fig3.png] |
+
+## Target Jurisdiction
+- [ ] CN (CNIPA — 国家知识产权局)
+- [ ] US (USPTO)
+- [ ] EP (EPO)
+- [ ] ALL
+
+## Patent Type
+- [ ] Invention patent (发明专利) — 20 year protection
+- [ ] Utility model (实用新型, CN only) — 10 year protection, apparatus claims only
+
+## Language
+- [ ] Chinese (中文)
+- [ ] English
+- [ ] Auto (detect from jurisdiction)
+
+## Claimable Subject Matter (可权利化的主题)
+[What types of claims do you expect?]
+- [ ] Method/process claims (方法权利要求)
+- [ ] System/apparatus claims (系统/装置权利要求)
+- [ ] Product claims (产品权利要求)
+- [ ] Computer-readable medium claims (计算机可读存储介质, US)
+- [ ] Other: [describe]
+
+## Known Related Patents/Papers
+[List any patents or papers you have already identified as relevant. Include patent numbers or DOIs.]
+
+| Reference | Type | Title | Why Relevant |
+|-----------|------|-------|-------------|
+| | Patent / Paper | | |
+
+## Filing Timeline
+- **Preferred filing date**: [date or ASAP]
+- **Priority claim**: [if claiming priority to earlier application]
+- **Publication deadline**: [if any]
+
+## Additional Notes
+[Any other information relevant to the patent application.]

package/templates/MANIFEST_TEMPLATE.md

@@ -0,0 +1,7 @@
+# Research Output Manifest
+
+> Auto-maintained by ARIS skills. Tracks all generated artifacts across the research lifecycle.
+> See [shared-references/output-manifest.md](../skills/shared-references/output-manifest.md) for the protocol.
+
+| Timestamp | Skill | File | Stage | Description |
+|-----------|-------|------|-------|-------------|

package/templates/NARRATIVE_REPORT_TEMPLATE.md

@@ -0,0 +1,49 @@
+# Narrative Report: [Your Paper Title]
+
+> **Template for Workflow 3 (`/paper-writing`).** Fill in all sections, then run `/paper-writing "NARRATIVE_REPORT.md"`.
+
+## Core Story
+[2-3 paragraphs: problem, approach, key finding, contribution]
+
+## Claims
+1. **[Main claim]**: [Statement with evidence]
+2. **[Supporting claim]**: [Statement with evidence]
+
+## Experiments
+
+### Setup
+- **Models**: [What models]
+- **Data**: [Datasets, splits]
+- **Hardware**: [GPUs, time]
+- **Baselines**: [Comparisons]
+
+### Experiment 1: [Name]
+[What you tested and why]
+
+| Method | Metric 1 | Metric 2 |
+|--------|----------|----------|
+| Baseline A | X.XX | X.XX |
+| **Ours** | **X.XX** | **X.XX** |
+
+**Interpretation**: [What the numbers mean]
+
+### Experiment 2: [Name]
+[Repeat pattern]
+
+## Figures
+1. **Figure 1**: [Type, what it shows, data source]
+2. **Table 1**: [Comparison, metrics]
+
+## Known Weaknesses
+- [Limitation 1]
+- [Limitation 2]
+
+## Related Work
+- **[Category 1]**: [Key papers, how you differ]
+- **[Category 2]**: [Key papers, how you differ]
+
+## Proposed Title
+[Title]
+
+## Target Venue
+[Venue] [Year]