resubmission-strategy
Turning a rejected grant into a funded one — triaging reviews, the introduction to resubmission, and strategic revision.
- License 许可
- MIT
- 最近更新
- 2026-11-07
- Stars
- ★ 0
- Forks
- 1
01 前言 Front Matter
- name
- resubmission-strategy
- category
- scientific
- description
- Turning a rejected grant into a funded one — triaging reviews, the introduction to resubmission, and strategic revision.
03 SKILL.md 正文
Overview
Most funded grants were once rejected: resubmission is the norm, not the exception. This skill covers triaging reviewer feedback (what to fix, what to rebut, what to ignore), writing the resubmission introduction (the 1-page "what changed" document), deciding how much to revise vs restructure, and the timeline and psychology of coming back stronger.
When to use
- Planning a resubmission after a grant rejection or low score
- Writing the NIH-style Introduction to Resubmission (or equivalent)
- Deciding whether to revise, restructure, or start over
- Responding to a "not discussed" (triaged) outcome
- Timing the resubmission: next cycle vs waiting
Core concepts
-
The score tells you the strategy: near-miss (just below payline) → targeted fixes; middling score → substantial revision of the weak sections; triaged/not discussed → fundamental restructuring or rethinking — possibly a different mechanism or aims.
-
Reviewer archetypes: the enthusiast (protect their support), the skeptic with fixable concerns (convert with data), the hostile/mistaken reviewer (rebut politely with evidence, and make the revision robust to their misreading), the vague reviewer ("more preliminary data" — interpret generously and deliver).
-
The introduction document: typically 1 page summarizing what changed in response to the major criticisms — this is what reviewers read first on resubmission; it must show responsiveness without groveling.
-
Fix, don't just argue: resubmissions that only rebut without new data or text changes rarely improve — reviewers want to see the proposal change, not just the response document.
-
The "new reviewer" problem: resubmissions often get new reviewers who never saw the old critiques — the proposal must stand alone, not depend on the introduction for context.
-
Diminishing returns: most mechanisms allow limited resubmissions (NIH: one resubmission as A1) — each round must be a serious upgrade; don't burn the resubmission on cosmetic changes.
-
Reading the summary statement forensically: distinguish fatal flaws (fix with new data) from misunderstandings (fix with clearer writing) from differences of opinion (address respectfully, don't capitulate) — each needs a different response.
-
The introduction page: a one-page (NIH) summary of what's changed — organized by reviewer concern, not chronologically; lead with the biggest changes, not the easiest.
-
New data vs new words: reviewers reward substantive changes — a major new experiment outweighs ten rewritten paragraphs; budget revision time accordingly.
Practical workflow
1. Triage the reviews (within a week, while fresh)
- Extract every criticism; sort into: fatal (must fix with new data/text), substantial (fix with revision), stylistic (fix cheaply), mistaken (rebut with evidence), out-of-scope (decline politely with rationale).
- Identify the pattern: do all reviewers hit the same weakness? That's your revision's center of gravity — fix it first and most visibly.
- Get an outside read: a colleague who hasn't seen the proposal reads the reviews cold — they'll spot which criticisms are valid and which are reviewer hobbyhorses.
2. Decide the revision scope
- Near-miss: keep the structure; strengthen weak aims, add the requested preliminary data, clarify the confusing passages. 4–8 weeks of work.
- Middling: restructure weak sections, possibly replace an aim, add substantial new data. 2–4 months.
- Triaged: fundamental rethink — new aims, new framing, possibly new mechanism (R21→R01, different institute) or a different funder. Don't resubmit the same proposal with new fonts.
- Sanity check: if the core idea scored poorly on significance/innovation (not just approach), revision can't fix it — rethink the idea.
3. Execute the revision
- Generate the new preliminary data first — it's the longest pole; text revision happens around it.
- Rewrite weak sections from scratch rather than patching — patched text reads patched.
- Mark changes clearly (many agencies want changed text indicated); keep a change log mapping each major criticism to its fix location.
4. Write the introduction
- Open by thanking reviewers and stating the revision's headline changes (3–4 bullets max).
- Address each major criticism: what was said, what you changed, where to find it — concise, factual, non-defensive.
- For disagreements: state your position with evidence, note the change you made anyway to address the underlying concern (this disarms).
- Close with why the proposal is now stronger — one confident paragraph.
5. Plan the resubmission timeline
- Work backward from the deadline: new experiments first (longest lead time), then analysis, then writing — most resubmissions fail on the experiment schedule, not the writing.
- Get fresh eyes before resubmitting: a colleague who hasn't seen the proposal reads the revision cold against the old critiques — they'll catch what you've gone blind to.
- Decide the venue question early: same study section (reviewers may remember) vs a different one (fresh start, new risks) — discuss with the program officer.
6. Quick-reference checklist
- [ ] Summary statement forensically categorized (fatal vs fixable vs opinion)
- [ ] Program officer consulted on scope and responsiveness
- [ ] New data addresses the major empirical critiques
- [ ] Introduction page leads with biggest changes, organized by concern
- [ ] Nothing resubmitted unchanged on the major critiques
- [ ] Strong sections protected; surgery focused on weak ones
- [ ] Fresh-eyes read against the old critiques before resubmission
- [ ] Venue decision made (same vs different study section)
Common pitfalls
- Cosmetic resubmission: new fonts, shuffled paragraphs, same weaknesses — reviewers remember, and scores drop.
- Arguing instead of fixing: a 10-page rebuttal with no new data — the introduction should describe changes, not litigate the old review.
- Ignoring the consensus criticism: all three reviewers flagged feasibility; you added a paragraph — fix the actual weakness with data or redesign.
- Overcorrecting to one reviewer: warping the proposal around a single hostile reviewer's hobbyhorse — weigh all reviews; the enthusiast's support matters too.
- Missing the window: waiting so long the preliminary data ages or the field moves — resubmit at the next viable cycle with real improvements.
- Resubmission fatigue: sending the A1 out of obligation rather than conviction — if you don't believe it's now fundable, don't submit; rethink instead.
- Resubmitting without addressing the major critique: hoping new reviewers won't notice — they often include the old reviewers, and the summary statement persists.
- Overhauling what worked: rewriting strong sections while under-fixing weak ones — protect what reviewers liked; operate surgically on what they didn't.
04 同仓库更多 Skills
web-scrape
Turn web pages into LLM-ready markdown via the Firecrawl API (renders JS, anti-bot handling). Use when browser_open/fetch fails on JS-heavy pages, or when you need clean markdown/structured extraction from a site. Requires FIRECRAWL_API_KEY.
design-taste-frontend
tasteskill v2, the anti-slop frontend skill. Art direction for websites, distilled from a study of 247 curated gallery sites (One Page Love, Awwwards Sites of the Day). Deals researched directions (idea, composition, type, palette, photographs, section blueprints) from a library, enforces a strict no-slop house style with a real-browser inspector, and runs a look-and-fix loop on screenshots. Use this whenever you build, design, redesign, restyle, polish or review a landing page, marketing site, portfolio, product page, one-pager, hero or any public-facing web page, in any stack (HTML/CSS, React, Next.js, Vue, Svelte, Astro, Tailwind) — also when the user says a page looks generic, templated, "AI-made", sloppy, or asks to make it look designed, premium, award-level or less like every other site. Not for dashboards or app screens.
test-driven-development
Use when implementing any feature or bugfix, before writing implementation code
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
requesting-code-review
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).