القائمة
Evaluates whether the paper's key insight is sufficiently novel by comparing with existing literature and analyzing novelty claims and differentiation from prior work.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Detect when the user wants to start a new CoPaper paper project or initialize a workspace, run `copaper --root . init` when needed, then inspect the current workflow phase and immediately begin the appropriate next phase work.
Inspect the current CoPaper workflow phase, move forward, move backward, or jump to a named phase when the user asks to view, advance, return, roll back, or switch phases in OpenCode, then invoke the skill for the active phase. Uses `copaper status --json` and `copaper set-phase` instead of editing state files.
Convert `paper.md` into a final LaTeX paper project using a user-provided conference or journal template placed under `templates/latex/`. Use when the user asks to generate the final LaTeX paper from the CoPaper markdown draft with a local template.
Helps create placeholder data (tables, figures, metrics) for evaluation sections with clear warnings that all numbers must be replaced with real experimental results. Also generates descriptions explaining what the data shows and how it supports the paper's claims and insights.
Detects undefined terms, unclear descriptions, and clarity issues in academic paper drafts to improve readability and understanding.
Teaches agents how to manage a CoPaper project through the OpenCode plugin tools, including dashboard, initialization, artifact readiness, workflow status, event log inspection, and phase updates. Use this skill whenever the task is about workflow/state management rather than direct paper writing.
| name | novelty-checker |
| description | Evaluates whether the paper's key insight is sufficiently novel by comparing with existing literature and analyzing novelty claims and differentiation from prior work. |
This skill evaluates the novelty and originality of a paper's key insight by comparing it with existing literature, analyzing novelty claims, and ensuring sufficient differentiation from prior work.
| File | Required | When to Read | Purpose |
|---|---|---|---|
paper.md | Required | Step 1 (start) | Primary analysis target — read insight and contribution claims |
relatedwork/summary.md | Optional | Step 2 | Literature overview for novelty comparison |
relatedwork/papers/*.md | Optional | Step 2 | Individual paper summaries for detailed comparison |
relatedwork/paper_list.bib | Optional | Step 2 | Bibliography for citation verification |
.agents/state.json | Tool gap | Final step | Checker-result persistence is not currently exposed by the OpenCode plugin; do not edit state directly in plugin-based workflows |
Do NOT read writingrules.md — the essential structure rules are inlined in the Paper Structure Reference section below.
You are an AI assistant performing novelty analysis of academic papers. Your comments are AI-generated and must be clearly marked as such. Your analysis should be:
| Marker | Meaning | When to Use |
|---|---|---|
<!-- AI Comments: | Start of AI-generated comment | ALWAYS use to begin every comment |
**AI-GENERATED NOVELTY ANALYSIS - FOR AUTHOR REVIEW** | Warning that content is AI-generated | ALWAYS include at the start of comment body |
[NOVELTY ISSUE TYPE] | Category of novelty problem | ALWAYS include to classify the issue |
[LOCATION] | Where the issue is found | ALWAYS include with section name and exact quote |
[SEVERITY] | How serious the issue is | ALWAYS include (Critical/Major/Minor) |
**END AI-GENERATED NOVELTY ANALYSIS** | End of AI analysis content | ALWAYS include before closing --> |
Definition: The core insight already exists in published literature with minimal or no difference.
| Type | Description | Example |
|---|---|---|
| Exact Duplicate | Identical insight published before | Claim: "We discovered X" → Same insight in prior paper |
| Near-Duplicate | Core idea same with minor variations | Prior work: "Use temporal patterns" → Current: "Use time-based patterns" |
| Rebranding | Same concept with different terminology | "Attack chain" vs "Kill chain" for same concept |
| Unaware of Prior Work | Insight exists but author unaware | Claims novelty for 5-year-old idea |
Definition: The insight is not sufficiently different from existing work to justify novelty claim.
| Type | Description | Example |
|---|---|---|
| Incremental Extension | Minor modification of existing work | Add one feature to existing method |
| Obvious Combination | Combining A + B in straightforward way | "Apply existing method M to domain D" |
| Parameter Tuning | Only differs in parameter choices | Same method with different hyperparameters |
| Scale Difference | Only difference is scale/dataset size | Same approach on larger dataset |
| Implementation Detail | Novelty claimed for implementation only | "We implemented X in Python" (original in C) |
Definition: Relevant prior work not cited or acknowledged.
| Type | Description | Example |
|---|---|---|
| Key Related Work Omitted | Important prior work not cited | Seminal paper on same topic missing |
| Recent Relevant Work | Recent similar work not cited | Paper from last 2 years with same insight |
| Cross-Domain Prior Art | Similar idea in different domain ignored | Same insight exists in NLP but paper is in security |
| Foundational Work Missing | Work that established the concept not cited | Original paper introducing the concept not cited |
Definition: Novelty claims exceed what is actually novel.
| Type | Description | Example |
|---|---|---|
| "First to..." Without Evidence | Claims first without verification | "First to use X for Y" but prior work exists |
| "Novel" Without Comparison | Claims novelty without comparing | "Novel approach" but no related work section |
| Broad Claims from Narrow Contribution | Over-generalizing novelty | "We revolutionize X" for minor improvement |
| Ignoring Limitations | Claims ignore known limitations | Claims novelty overcame challenge that was already solved |
Definition: Arguments for why the work differs from prior art are weak or unconvincing.
| Type | Description | Example |
|---|---|---|
| Vague Differences | "Different from X" without specifics | "Our method differs from prior work" - how? |
| Superficial Comparison | Comparison misses key aspects | Compares only on one metric, ignores others |
| Strawman Comparison | Compares to weak/outdated baselines | Claims improvement over 10-year-old method |
| Missing Quantitative Differentiation | No numbers to show difference | "Better than X" without metrics |
| Unfair Comparison | Compares favorable aspects only | Compares speed but not accuracy |
Definition: Paper doesn't clearly state in what dimension(s) it is novel.
| Type | Description | Example |
|---|---|---|
| Novelty Type Unclear | Not stated if novelty is problem, method, or insight | "Novel contribution" - what kind? |
| No Novelty Positioning | Doesn't position relative to prior work | No statement of what's new vs. what's adopted |
| Contribution Not Isolated | Can't identify what's novel vs. standard | Paper mixes known and novel without distinction |
Types of Novelty (authors should clarify which apply):
Definition: Similar insights exist in other domains but not acknowledged or adapted.
| Type | Description | Example |
|---|---|---|
| Direct Applicability | Method from domain A directly applies to B | Claims novelty but same method used in other domain |
| Conceptual Similarity | Same concept exists elsewhere | "Temporal patterns" known in speech, applied to security |
| Technique Transfer | Known technique, new domain | Standard ML technique, new application |
| Opportunity for Adaptation | Could strengthen paper by acknowledging | Could cite cross-domain work and explain adaptation |
Definition: The novelty claim doesn't align with what related work section shows.
| Type | Description | Example |
|---|---|---|
| Contradictory Claims | Novelty claim contradicts related work | Claims "first" but related work cites similar approach |
| Missing Gap Analysis | No clear gap identified | Related work described but no gap stated |
| Gap Not Filled | Gap identified but not addressed | Identifies gap G but contribution doesn't address G |
| Related Work Incomplete | Missing categories of related work | Only cites 5 papers when field has 50+ |
IMPORTANT: All comments generated by this skill are AI-generated analysis and suggestions. They must be clearly marked with "AI Comments:" to distinguish them from human reviewer feedback.
All Novelty Check Comments must follow this standardized format:
<!-- AI Comments:
**AI-GENERATED NOVELTY ANALYSIS - FOR AUTHOR REVIEW**
[NOVELTY ISSUE TYPE]
<Type from the 8 categories above>
[LOCATION]
Section: <section name>
Text: "<exact quote of the problematic text>"
[PROBLEM DESCRIPTION]
<explanation of why this is a novelty problem>
[DETECTED ISSUE]
<specific description of the novelty concern>
[EXISTING WORK]
<relevant prior work that challenges novelty claim>
- Paper: <citation or description>
- Similarity: <what is similar>
- Key difference (if any): <what differs>
[IMPACT ON NOVELTY]
<how this affects the novelty claim>
- Strength of prior work: <how strong/published/recognized>
- Overlap: <percentage or extent of overlap>
- Differentiation needed: <what would be needed to differentiate>
[SUGGESTED ACTIONS]
<concrete suggestions on how to address>
1. <cite and acknowledge>
2. <differentiate explicitly>
3. <strengthen novelty argument>
4. <adjust claim scope>
[SEVERITY]
Critical / Major / Minor
- Critical: Novelty claim is undermined by existing work
- Major: Significant weakness in novelty argument
- Minor: Could strengthen novelty positioning
**END AI-GENERATED NOVELTY ANALYSIS**
-->
From paper.md:
From relatedwork/ folder (if exists):
summary.md and individual paper summariespaper_list.bib for comprehensive literatureIf relatedwork folder is incomplete, search for:
For each relevant prior work:
For each novelty claim in the paper:
For each similarity found:
For each novelty issue found:
Generate a comprehensive novelty assessment.
The paper follows CoPaper structure. Key rules for this checker:
# to #####): Structural headings only — no body text allowed under these.######): Content paragraphs. Title = topic sentence (≤50 chars). Body = supporting text (≤500 chars).<!-- description: ... --> guide what each section should contain.After analyzing the paper, provide a summary:
## Novelty Assessment Summary
**Paper**: [Paper title]
**Key Insight**: [Extracted core insight]
### Novelty Evaluation
**Overall Novelty Strength**: Strong / Moderate / Weak
**Novelty Dimensions Claimed**:
- [ ] Problem Novelty
- [ ] Method Novelty
- [ ] Insight Novelty
- [ ] Theoretical Novelty
- [ ] Empirical Novelty
- [ ] Systems Novelty
### Prior Work Comparison
**Directly Similar Work** (high overlap):
1. [Paper A] - [Similarity description] - [Key difference]
**Related Work** (partial overlap):
1. [Paper B] - [Similarity description] - [Key difference]
**Cross-Domain Similarities**:
1. [Paper C from domain X] - [Similarity description]
### Novelty Issues Found
- Critical: X
- Major: Y
- Minor: Z
### Strengths
- [What novelty aspects are strong]
### Weaknesses
- [What novelty aspects need strengthening]
### Recommendations
**To Strengthen Novelty**:
1. [Specific recommendation]
2. [Specific recommendation]
**To Adjust Claims**:
1. [What claims should be modified]
**To Add Citations**:
1. [Missing citations that should be added]
### Risk Assessment
**Rejection Risk**: High / Medium / Low
- [Reasoning]
**Reviewer Concerns Likely**:
1. [Specific concerns reviewers might raise]
Insert detailed HTML comments at problematic locations following the comment structure defined above. All comments must:
<!-- AI Comments:**AI-GENERATED NOVELTY ANALYSIS - FOR AUTHOR REVIEW****END AI-GENERATED NOVELTY ANALYSIS** before the closing -->For detailed comment examples, see examples.md in this skill directory.
Whenever the paper claims:
Verification needed:
For incremental improvements:
Assessment criteria:
When checking method novelty:
Questions to ask:
When analyzing the gap:
For cross-domain work:
All comments are AI-generated: Every comment inserted by this skill is generated by AI analysis and must be clearly marked with "AI Comments:". These are NOT human reviewer feedback and should not be treated as such.
Be honest about novelty: Not every paper needs groundbreaking novelty. Incremental contributions are valid but should be positioned appropriately.
Consider venue expectations: Top-tier venues (NSDI, OSDI, CCS, SOSP) expect strong novelty. Workshops have lower bars.
Novelty is not binary: It's a spectrum. Help authors understand WHERE they are on that spectrum.
Prior art is not bad: Citing and differentiating from prior art STRENGTHENS the paper, not weakens it.
Check relatedwork folder: If it exists, use it. If incomplete, note that additional literature search is needed.
Authors should verify: AI may not find all prior art. Authors should do their own thorough search.
Multiple novelty dimensions: A paper can be novel in multiple ways. Help identify all applicable dimensions.
Avoid discouraging: Even if novelty is weak, provide constructive paths forward.
Publication risk assessment: Provide realistic assessment of publication risks at target venues.
Use this checklist during analysis: