Skip to content

K-Dense-AI/scientific-agent-skills / skills/research-grants

research-grants

Verified

Supports research proposal preparation and review for NSF, NIH, DOE, DARPA, and Taiwan NSTC, including opportunity-specific requirements, aims, review criteria, budgets, broader impacts, forms, and resubmissions. Use for investigator-authored grant development, compliance matrices, and proposal critiques.

references/nsf_guidelines.md

View on GitHub
# NSF proposal guidance

Reviewed 2026-10-01. Use the solicitation, amendments, and effective policy notices
before the base PAPPG. The [current PAPPG index](https://www.nsf.gov/policies/pappg)
still identifies NSF 24-1; draft replacement guidance is not an effective rule.
[Supplement 1](https://www.nsf.gov/policies/document/pappg24-1-supplement-1)
and [Supplement 2](https://www.nsf.gov/policies/document/pappg24-1-supplement-2)
modify it. Record which rules apply to the proposal and prospective award.

## Start from the opportunity

Record its NSF number, revision, due date and time zone, eligible organizations and
investigators, proposal type, budget, duration, and required documents. Use
[funding search](https://www.nsf.gov/funding/opportunities) for current opportunities;
a funded award or an archived solicitation does not establish an open call.
Use current directorate and program names from the opportunity, not an old org chart.

## Review and narrative

[Intellectual Merit and Broader Impacts](https://www.nsf.gov/policies/pappg/24-1/ch-3-proposal-processing-review)
are both required and receive full consideration. NSF does not prescribe a universal
50/50 numerical weighting. Include any additional solicitation criteria.

For intellectual merit, explain the question, existing evidence, specific gap,
new contribution, design, analysis, uncertainty, alternatives, and available expertise.
For broader impacts, connect an achievable benefit to activities, resources, owners,
and assessment. Research itself can deliver broader impacts; outreach is not mandatory
for every project. See [broader_impacts.md](broader_impacts.md) for drafting examples.

## Application package

Under [PAPPG Chapter II](https://www.nsf.gov/policies/pappg/24-1/ch-2-proposal-preparation),
standard research proposals generally use the following package. Special proposal
classes and solicitations can differ; construct the actual checklist from those rules.

| Component | Check |
| --- | --- |
| Project Summary | Up to one page; identify Overview, Intellectual Merit, and Broader Impacts separately. |
| Project Description | Normally up to 15 pages. Include a separate **Broader Impacts** heading on its own line, even if impacts also appear elsewhere. |
| Prior NSF support | When applicable, include Results from Prior NSF Support within the description, up to five pages; check the current-award and preceding-five-year triggers for each PI/co-PI, and separately identify intellectual merit and broader impacts. |
| References | Complete identifiable citations; no narrative overflow. |
| Budget and justification | Match the narrative, person-months, and subawards; normally five pages of justification per proposing organization. |
| Facilities/resources | Describe available resources without quantifying voluntary committed cost sharing. |
| Senior/key documents | Certified SciENcv biosketch and Current and Pending (Other) Support, COA information, and separate Synergistic Activities document. |
| Mentoring | One-page plan when supporting graduate students or postdoctoral scholars, including supported people at subrecipients and collaborative organizations. |

The NSF Common Form biosketch has no overall page limit; individual fields/products
have limits. Synergistic Activities is a separate one-page document with up to five
examples, not a section in an old three-page biosketch. Use the current SciENcv export
and required certification; do not manufacture a lookalike or edit a certified PDF.
Check current research-security training and disclosure certifications under
Supplement 1 and the submission system.

### Data management and sharing

Supplement 2 specifies the **Research.gov DMSP tool**, released April 27, 2026.
The earlier two-page PDF-upload workflow applies to proposals submitted before that
date. Prepare the plan using the current tool and NSF-unit guidance. Address data,
repositories, timing, access, reuse, preservation, and justified restrictions. One
unified plan covers a simultaneously submitted collaboration and subawards. Under
the supplemented policy, data supporting publications are shared at publication
unless an exception is justified. Confirm the award's applicable public-access terms.

### Formatting

For ordinary uploaded proposal text, Chapter II permits Arial (not Arial Narrow),
Courier New, or Palatino Linotype at 10 points or larger; Times New Roman or Computer
Modern at 11 points or larger. Use margins of at least one inch, at most six lines per
vertical inch, and no larger than US letter paper. Smaller text is permitted for
specified equations/figures/tables/captions but must remain legible. Check exported
PDFs at actual size. Follow Research.gov pagination instructions and auto-checks.

## Budget decisions

- NSF's normal two-month salary policy applies per senior/key person **across all NSF
  grants**, not just summer salary on one proposal. Explain anticipated excess in
  the budget justification and obtain the required approval through the award.
- Do not offer voluntary committed cost sharing to improve competitiveness. Include
  mandatory sharing only when the solicitation requires it.
- Use the organization's applicable negotiated indirect-cost rate and base, current
  award terms, and governing cost rules. Equipment thresholds and subaward portions
  included in MTDC are not safely inferred from the unamended 2024 text; Chapter II
  states that 2 CFR takes precedence where inconsistent. Have sponsored research
  confirm the effective capitalization policy and rate agreement.
- Separate employees doing project work from eligible participants in a training
  activity; not every student payment is participant support. Classify tuition,
  fringe, and indirect-cost exclusions explicitly.

## Special proposal types

[CAREER NSF 22-586](https://www.nsf.gov/funding/opportunities/career-faculty-early-career-development-program/nsf22-586/solicitation)
requires an eligible untenured assistant professor or equivalent with at least a
50% tenure-track/equivalent appointment and no prior CAREER award, among other
conditions. There is **no general six-years-since-PhD rule**. Check the three-competition
limit, one proposal per competition, no co-PIs, departmental letter, five-year plan,
and directorate-specific minimum award amount. A minimum is not a budget ceiling.
Read the current deadline from the program listing rather than inferring it from July.

PAPPG RAPID and EAGER are special procedures requiring prior program-officer
consultation/authorization and their specified concept outlines. Check the current
PAPPG for limits (RAPID normally $200,000/one year; EAGER $300,000/two years, both including indirect costs), review
exceptions, and proposal-specific instructions. Do not treat them as ordinary open
calls or a guarantee of no external review.

Collaborations may use a single proposal with subawards or separately submitted,
linked proposals. Follow the selected method's shared-document and lead-organization
instructions rather than uploading duplicate near-identical narratives everywhere.

## Submission and revision

Use the opportunity's permitted route, normally Research.gov or the supported
Grants.gov route. FastLane is not the current proposal-submission workflow. Confirm
institutional registrations, PI affiliation, and AOR submission authority early.
Deadline times are normally 5 p.m. submitting organization's local time; the
solicitation and applicable exceptions control. Inspect the assembled application and
agency receipt/validation status; a local upload is not submission confirmation.

NSF generally treats a revised, previously declined proposal as a new proposal, with
no NIH-style A1 introduction. Address the reviews in the scientific plan, check any
program resubmission restrictions, and use permitted program-officer feedback.
Reviews and historical funding statistics do not guarantee a future award or timeline.
add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution.Stars: 115kUpdated Claude CodeCodexVerifiedInstall
add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation.Stars: 115kUpdated Claude CodeCodexVerifiedInstall
diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope.Stars: 115kUpdated Claude CodeCodexVerifiedInstall
fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks.Stars: 115kUpdated Claude CodeCodexVerifiedInstall