Oral Paper Skill
A skill that checks a manuscript against practices of 883 ICLR, ICML and NeurIPS Oral papers and suggests edits
- 550 ★ on the repository
- Last commit
Minimal access
Why this level
- The skill consists of instructions and reference files and runs nothing by itself.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/oral-paper-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill consists of instructions and reference files and runs nothing by itself.
This is third-party code. Review the repository files before installing.
What it does
Oral Paper Skill helps the author of a research paper learn from works picked for oral presentation at ICLR, ICML and NeurIPS. The author supplies an idea, a draft or an experiment plan. The agent selects fitting examples and shows how they frame a contribution, build evidence and choose comparisons. The skill runs in two modes. In the first it compares a manuscript with the examples and points to improvements: every piece of advice has a source, a reason, the state of the manuscript and a concrete edit. In the second it explains good practices and guides a review of your own work. The knowledge comes from 883 Oral papers: ICLR 2025–2026 (436), ICML 2025–2026 (288) and NeurIPS 2024–2025 (159). Seven practices are distilled: research tension, contribution delta, evidence matched to the claim, meaningful comparisons, conditions beside conclusions, what a resource enables and a bounded lesson. By default the agent picks only the relevant practices and gives up to three edits.
Who it is for. PhD students and researchers preparing a paper for ICLR, ICML or NeurIPS who want to check the introduction, contribution and experiments before submission.
Good fit when
- The introduction does not say how the contribution differs from prior work and you want to find the weak spot
- A claimed cost win needs comparisons and it is unclear which resource unit to measure
- You need to plan the main figure and key experiments around the claimed contribution
- After a rejection you want a review of how strong similar papers build their evidence
- You want to study on examples how Oral papers state the limits of a conclusion
Not a fit when
- You want an acceptance forecast: the skill does not predict committee decisions or know venue criteria
- You need grammar fixes or reference formatting: other tools cover that
- You need advice on figures and proofs from full texts: the base is built from abstracts
- The manuscript covers a niche topic outside machine learning with no examples in the base
Example request
Use $oral-paper-skill to compare my paper's introduction with fitting Oral papers and name the three most important edits with sources.Limitations
- The base is built from abstracts of 883 papers, not full texts. Advice on figures, experiment details and proofs needs the paper itself.
- The author states that the effect of using it has not been checked by an independent comparison.
- Oral status is a reason to pick an example and not an official standard of the venue.
- For the agent to find and read the cited papers it needs a tool with web or file access.
- The last commit in the repository dates from September 2026.
How to disable. Delete the `oral-paper-skill` folder from `~/.claude/skills/` or `~/.codex/skills/`.
FAQ
How do I install Oral Paper Skill?
Clone the repository and put the `skills/oral-paper-skill` folder into the agent's skills directory, for example `~/.claude/skills/` or `~/.codex/skills/`. Call the skill by name: `$oral-paper-skill`.
Is there a way to use it without installing the skill?
Yes, the repository holds short prompts in Chinese and English: `prompts/quick-prompt-en.txt`. Paste the text into a chat and attach your idea or manuscript.
Which venues does it cover?
ICLR 2025–2026, ICML 2025–2026 and NeurIPS 2024–2025: 883 Oral papers in total, checked against the official lists on 13 September 2026.
Will it tell me whether a paper gets accepted?
No. The skill does not predict committee decisions and does not replace peer review. It gives edits for contribution clarity and the link between evidence and claim.
How much advice does it give at once?
By default no more than three edits or one review exercise. It takes only the relevant practices out of seven and does not run the paper through all of them.
Discussion
No questions or reviews yet.
Related
SkillsA skill for writing and revising ML, CV and NLP papers: section structure, paragraph flow and reviewer-style self-review
Skill with 30 prompt templates for research papers in English and Chinese, from literature review to grant proposals
paper-reading-skill turns one arXiv AI or robotics paper into a report with each claim tied to evidence in the source.
A skill set for academic papers: deep method analysis, visual slides and a coherent write-up from a paper PDF
A Claude Code skill for research papers: a five-stage pipeline, section rules, a style audit and an independent red-team review.
Skill for writing economics papers, drawing on advice from more than 50 authoritative guides
README in short
The README covers the project goal, two modes, seven practices, a quick start with a prompt or install and the sources of 883 papers. It also states the limits: abstract-level evidence and no independent check of the effect.
README badge
Original SKILL.md
--- name: oral-paper-skill description: Help authors learn from exemplary ICLR, ICML, and NeurIPS papers through source-linked manuscript comparisons, concrete writing and experiment suggestions, and guided reflection. Use for research storytelling, contribution framing, paper comparison, figure planning, or research retrospectives. Not an acceptance predictor or a routine grammar/citation formatter. --- # Oral Paper Skill Turn useful practices from exemplary papers into concrete revisions and focused learning. Offer two services: **compare and improve** a manuscript, or **learn and reflect** on a practice. Preserve the author's research purpose and judgment. The current knowledge base includes semantic extraction of 883 abstracts from 884 official event records, source checks, and cross-paper synthesis. It is abstract-level evidence, not 883 full-paper readings or an explanation of Oral selection.