rebuttal-response-skills
Use when planning, drafting, revising, compressing, or auditing conference rebuttals and author responses in OpenReview Markdown or PDF. Supports exact reviewer mapping, evidence and experiment integration, guided risk decisions, venue-format validation, and blind reviewer/AC audits.
sidiangongyuan · 已有静态检查记录
安装与使用
运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。
npx skills add sidiangongyuan/codex-skills-library --skill rebuttal-response-skills技能说明
name: rebuttal-response-skills description: Use when planning, drafting, revising, compressing, or auditing conference rebuttals and author responses in OpenReview Markdown or PDF. Supports exact reviewer mapping, evidence and experiment integration, guided risk decisions, venue-format validation, and blind reviewer/AC audits. license: MIT
Rebuttal Response Skills
Goal
Produce reviewer-specific responses that are direct, easy to verify, and strictly grounded in the submitted paper or approved rebuttal evidence. Keep internal risk control rigorous without turning the visible response into a stream of disclaimers.
Choose the Workflow
Use the full workflow for a complete rebuttal, multiple reviewers, new evidence or experiments, a meta-review, or a final acceptance audit. Maintain four internal artifact classes in the task's rebuttal directory:
- Exact review archive, such as
reviews.md. - Concern/evidence map with a shared decision ledger, such as
concern_map.md. - One response per reviewer.
- Final audit record, such as
rebuttal_audit.md.
Use the lightweight workflow for a bounded sentence rewrite, one isolated question, or minor formatting. Edit inline and create no files unless the user asks. Preserve the original technical meaning and claim strength by default. The supplied sentence is sufficient for a pure prose edit; ask for context only when clarity would require changing a factual claim, number, quotation, or scope statement.
Read references/response-patterns.md when drafting a full response, integrating experiments, resolving a risk marker, running a blind audit, or performing final format verification.
Establish Sources of Truth
- Treat the complete review text supplied by the user as the review source of truth. If the user later supplies a revised version, use it and record the replacement. Check the platform only when text is incomplete, versions conflict, or the user asks. Do not draft a full response from a review summary; request the exact weaknesses, questions, score, and confidence.
- Read the submitted paper and supplement as the reviewer saw them. Prefer the submitted PDF for claims and values; use source files to locate content only when their submitted status is verified.
- For rebuttal-only work, treat the submission as immutable evidence. Record its hash or version before work and verify it is unchanged at the end. Edit it only under a separate explicit request.
- Determine the official response mode, per-response or shared limit, anonymity rules, Markdown/LaTeX support, and link or attachment policy. Verify current official instructions when the venue is known and the user has not supplied them.
Map Every Concern
Read each review sentence by sentence before drafting. For every substantive point, record:
- exact original sentence and source section;
- premise and actual concern;
- explicit request or score-shifting condition;
- submitted and rebuttal evidence, with exact source locations;
- evidence gap and proposed response mode;
- one coverage status:
fully answered,partially answered,untested boundary, orpending evidence.
Keep a separate map per reviewer and a shared decision ledger for user-confirmed values, prohibited wording, excluded experiments, cross-reviewer terminology, and stop decisions.
- Preserve the review's visible order unless the user explicitly requests a different order. A meta-review may change emphasis, not sequence.
- Match labels to the source. Use
W1,Q1, and similar labels rather than genericC1labels. - Prefer the reviewer's complete decisive sentence or sentences as the heading. Shorten only for precision or a binding format limit.
- Never place a paraphrase or stitched fragment inside quotation marks.
- Merge true duplicates internally. Preserve separate headings when overlapping questions contain distinct requests.
- Every substantive point must be answered, explicitly combined with a duplicate, marked as an untested boundary, or held for one necessary user decision.
Build the Response
Start each reviewer response with at most two short sentences that state the answer order. Restore method context only when the answer would otherwise be hard to understand; do not mechanically re-explain the paper.
For every issue:
- Give the direct answer in the first sentence.
- Explain only the mechanism, definition, or setup needed for that issue.
- Present evidence before interpretation.
- State the narrow implication supported by that evidence.
- Add a concrete camera-ready clarification only when the submitted wording actually needs correction.
Writing rules:
- Prefer short declarative sentences and terminology already used in the submission. Introduce a new term only when necessary and define it once.
- Use the order direct answer, evidence, and supported impact. Avoid opening or closing with an apology, “we acknowledge,” or a pre-emptive concession when the evidence is already clear.
- When a table fully reports a secondary metric or tradeoff, point to the measured value and explain its impact only if it changes the reviewer's question, the comparison's fairness, or the claim's applicability. Do not manufacture a negative rhetorical ending from a table-visible value.
- If the review relies on an incorrect or incomplete premise, correct it politely with submitted evidence, then answer the underlying concern.
- Stay within that reviewer's explicit concerns. Shared evidence may be reused, but do not import another reviewer's limitations, baselines, or side issues.
- Acknowledge an especially useful or decision-critical question once when it improves the exchange. Avoid routine praise and repeated gratitude.
- Use a closing contribution synthesis only for a low-score review or when it is needed to change the overall assessment.
- Never promise an experim