# Pods Author Response > Use when drafting ACM PODS author responses, covering the short ~48-hour double-anonymous rebuttal that corrects misreadings of proofs and definitions, and — distinctively — the shepherded revision cover letter that maps every required item to a concrete change in the resubmitted paper and survives the shepherd's second read. - Skill: `thedixitjain/pods-author-response` (Agent Skill) - Install (CLI): `npx skillmds add thedixitjain/pods-author-response` - Raw SKILL.md: https://api.skillmd.com/api/skills/thedixitjain/pods-author-response/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: thedixitjain (https://skillmd.com/u/thedixitjain) - Updated: 2026-09-08 - Page: https://skillmd.com/skills/thedixitjain/pods-author-response --- # PODS Author Response Use this after PODS reviews are released. PODS has **two distinct speaking turns**, and conflating them is a common mistake: a short **rebuttal** (about 48 hours, a few thousand characters) right after reviews, and — if you receive a **revision** decision — a **revision cover letter** accompanying a resubmitted paper that a **shepherd** re-reads. Both stay double-anonymous: do not reveal authors, affiliations, or a named system, even to strengthen a point. ## Triage (both turns) - Answer what affects correctness first: a claimed error in a proof, an unstated assumption, a mismatch between an upper and a lower bound, or an out-of-scope objection. - Distinguish a **misreading** (the reviewer misunderstood a definition or missed an existing appendix proof — fixable now, in the rebuttal) from a **real gap** (a step that genuinely needs to be written — fixable only in the revision). - Keep every word anonymous: no institution, no funder, no real system name, no "our earlier work" phrased in the first person. ## Turn 1 — the 48-hour rebuttal Short and correctness-focused. The rebuttal exists to **clarify misunderstandings and factual errors via pointers to the submitted paper** — not to present new mathematics. - Point to the exact definition, lemma, or appendix proof the reviewer overlooked ("the proof the reviewer requests is already in Appendix B, Lemma 7"). - Concede a genuine error plainly and state whether it is fixable in a revision. - Do **not** paste a new proof the reviewers cannot verify in 48 hours; promising an unwritten proof does not move a theory review — it flags the gap. ## Turn 2 — the revision cover letter (the distinctive PODS move) A revision (minor or major) is a genuine shepherded second round. The cover letter is a **change ledger**: for every required item, either make the change and show exactly where, or explain precisely why it is unnecessary — and satisfy the shepherd. ```text [R1.1] Required: "the complexity bound in Thm 3 is not justified" -> Action: DONE | added the full accounting in §4.2 and the deferred steps in Appendix B -> Where: §4.2 (para 2), Appendix B (Lemma 7, revised) [R1.2] Required: "state the assumption the lower bound rests on" -> Action: DONE | Thm 4 now labeled as conditional on OMv; stated at point of use (§5) [R2.1] Suggested: "extend the dichotomy to self-joins" -> Action: DECLINED (with reason) | self-joins need a different reduction; out of scope for the revision window, now stated as an explicit open problem (§7) rather than implied ``` The rule that turns a revision into an acceptance: **address every required item.** An item neither done nor explicitly and defensibly declined is what the shepherd's read punishes. ## Reviewer pushback patterns | Pushback | What it signals | PODS-ready response | |---|---|---| | "The proof of Thm X has a gap" | Correctness doubt | Fix the step in the revision and show it; if it is a misreading, point to the existing proof in the rebuttal | | "The lower bound does not match the upper bound" | Optimality unproven | Close the gap, or scope the optimality claim to the model actually proved | | "This assumption is hidden/unstated" | Rigor gap | Label the conditional result at its point of use; state the assumption explicitly | | "The model trivializes the result" | Modeling objection | Defend the model as realistic, or adjust it and re-derive; do not ignore | | "Overlaps prior result Y" | Novelty/delta doubt | Sharpen the theorem-level delta; add the missing comparison of results | ## Anonymity in the response (easy to slip) - Refer to your own prior work in the third person, as in the paper. - Describe revisions without linking to an identifying arXiv preprint or homepage; reference the in-PDF appendix location instead. - Do not thank a named collaborator, advisor, or funder inside the rebuttal or cover letter. ## Calibration - Respond to the criterion the reviewer raised (correctness, tightness, modeling, novelty), not the one you would rather defend. - Length and format norms for both turns vary by cycle; confirm the current instructions before sending. - The shepherd decides the revised paper; make the **paper**, not the letter, carry the corrected mathematics. ## Output format ```text [Turn] 48-hour rebuttal / revision cover letter [Priority issue] [Decision dimension] correctness / tightness / modeling / significance / clarity [Change ledger] DONE/PARTIAL/DECLINED + where in paper/appendix> [Anonymity check] ``` --- **Source:** [`brycewang-stanford/Awesome-Journal-Skills`](https://github.com/brycewang-stanford/Awesome-Journal-Skills) → `PODS-Skills/skills/pods-author-response/SKILL.md`