| name | grants-writer |
| description | Grant proposal writing: research grants, innovation grants, government funding, and charitable foundation applications — structure, narrative, budget justification, and evaluation criteria |
Grants Writer Skill
When to activate
- Writing a research grant proposal (NSF, NIH, UKRI, EU Horizon, etc.)
- Applying for an innovation or startup grant (SBIR, Innovate UK, regional schemes)
- Writing a foundation grant application (charitable, philanthropic)
- Structuring a grant budget and justification narrative
- Reviewing a draft proposal against the funder's evaluation criteria
- Writing a project abstract or executive summary for a grant application
When NOT to use
- Investor pitch decks — use the pitch-deck skill
- Academic paper writing — use the lit-review skill
- Contract bids or RFP responses — different format and procurement rules
- Fundraising for a non-profit via donations — different audience and strategy
Instructions
Research grant structure
Write a research grant proposal for [funding body].
Funding body: [NSF / NIH / UKRI / EU Horizon / other]
Grant scheme: [scheme name and call]
Budget: $[X] over [X] years
Discipline: [describe research field]
Research question: [what you want to investigate]
Team: [PI, Co-Is, institution]
Deadline: [date]
Proposal structure (adapt to funder's specific format):
1. PROJECT SUMMARY / ABSTRACT (1 page, often public):
- Background: why this problem matters
- Objective: what the research will do
- Methods: how you will do it (brief)
- Expected outcomes: what you will produce
Rule: write this last, after the full proposal is complete
2. INTRODUCTION / SIGNIFICANCE (1-2 pages):
- The problem: what knowledge gap exists?
- Why it matters: scientific or societal importance
- Why now: what makes this the right time to fund this work?
- Prior work: what has been done, and what's left to do?
[Cite extensively — demonstrates literature mastery]
3. OBJECTIVES AND AIMS (most important section):
Format: 3 specific aims, each independent but contributing to the overall goal
Aim 1: [Test hypothesis / develop method / demonstrate X]
Aim 2: [...]
Aim 3: [...]
Rule: aims should be achievable even if one fails. Reviewers penalise "if Aim 1 fails, Aims 2 and 3 cannot proceed."
4. RESEARCH PLAN / METHODS:
For each aim:
- Rationale: why this approach?
- Methods: exactly how
- Expected results: what you anticipate finding
- Potential problems and alternatives: how you'll handle failures
- Timeline: milestones per quarter
5. INNOVATION:
[What is genuinely new about this work?]
[How does it go beyond the current state of the art?]
[What new tools, methods, or knowledge will it produce?]
6. TEAM AND ENVIRONMENT:
[PI qualifications — relevant publications and experience]
[Collaborators and their specific roles]
[Institution resources that support this project]
7. BROADER IMPACTS (NSF-specific) / PUBLIC ENGAGEMENT (UKRI):
[Who benefits and how?]
[Training: how does this fund student/early-career researchers?]
[Outreach: how will findings reach beyond academia?]
Write the proposal sections for my specific grant and research question.
Innovation / startup grant
Write a grant application for [innovation scheme].
Scheme: [SBIR Phase I / Innovate UK Smart Grant / EU EIC Accelerator / regional scheme]
Organisation: [company name, size, stage]
Technology: [describe the innovation]
Problem: [market problem being solved]
Grant amount: [X]
Match funding required: [yes — X% / no]
Innovation grant structure:
1. EXECUTIVE SUMMARY (1 page):
- The problem and market opportunity
- Your solution and what makes it novel
- What the grant will achieve
- Team overview
- Grant ask and match funding
2. PROBLEM STATEMENT:
- Market pain: who has this problem and how severely?
- Current solutions and their limitations
- Market size: TAM, SAM (bottom-up, evidence-based)
- Urgency: why solve this now?
3. SOLUTION AND INNOVATION:
- What you've built or are building
- The novel element (technical / business model / process)
- Why this is hard to replicate (barriers to entry)
- IP position: patents, trade secrets, data advantages
4. TECHNICAL WORK PLAN:
Funder wants to know grant money is used for credible R&D, not just marketing.
Work packages: [WP1: X / WP2: Y / WP3: Z]
For each: objective, activities, deliverables, timeline, lead person
Technical risks and mitigation:
| Risk | Likelihood | Impact | Mitigation |
|---|---|---|---|
5. COMMERCIALISATION AND MARKET:
- Business model: how will this generate revenue?
- Go-to-market: first customers, channel strategy
- Traction: pilots, LOIs, letters of support
- Financial projections (3 years post-grant)
6. TEAM:
- Why this team for this problem?
- Key hires planned with grant funding
- Advisory board / strategic partners
7. BUDGET:
- Personnel: [X]
- Equipment / software: [X]
- Subcontractors: [X]
- Overheads: [X%]
- Total: [X]
[Every line item justified in narrative]
Write the innovation grant application for my organisation and scheme.
Budget justification
Write a grant budget justification for [proposal].
Total budget: $[X]
Budget period: [X years]
Categories: [personnel / equipment / travel / indirect costs / subcontracts]
Budget justification narrative (one paragraph per category):
PERSONNEL:
"[PI name] (20% effort, Year 1-3): [PI] will provide overall scientific direction, lead [specific activities], and ensure timely completion of project milestones. [20% of annual salary + fringe = $X].
[Postdoc name] (100% effort, Year 1-2): [Postdoc] will conduct [specific experiments/analyses] as described in Aims [X]. [Annual salary + fringe = $X]. A second postdoc will be recruited in Year 2 to lead [Aim Y] work. [$X]."
EQUIPMENT:
"[Equipment name] ($X): Required for [specific experimental task described in methods]. The [institution] does not have access to [equipment], and purchase is more cost-effective than rental/core facility use because [reason]."
TRAVEL:
"Domestic travel ($X/year): To present findings at [conference names] and attend project team meetings. International travel ($X): To [conference] in Year 2 to disseminate [specific milestone results]."
INDIRECT COSTS:
"Indirect costs are calculated at [X]% of modified total direct costs (MTDC) per [Institution's] federally negotiated rate agreement dated [date]."
SUBCONTRACTS:
"[Partner institution] will provide [specific services/expertise] not available at [home institution]. Subcontract budget: $X for [X person-months of named PI at partner]."
Write the budget justification for my specific grant categories and amounts.
Reviewer criteria alignment
Align my grant proposal to reviewer criteria for [funding scheme].
Funding scheme: [name]
Evaluation criteria: [paste from funder's guidance or describe]
My proposal draft: [paste or describe key elements]
Alignment check:
For each criterion, confirm:
Criterion: [e.g. "Scientific merit and novelty"]
Weight: [X% of overall score]
What reviewers look for: [describe]
My proposal's response: [section + specific language]
Gap: [what's missing or weak]
Improvement: [specific change to make]
Common criteria and how to address them:
SIGNIFICANCE / IMPACT:
- Show the problem is important with data (not opinion)
- Quantify: "X million people affected" / "$Y billion market" / "[Z]% of patients lack effective treatment"
- Explain why solving this matters beyond the research community
INNOVATION:
- Reviewers want to see what's genuinely new, not incremental
- Use: "Unlike existing approaches that [limitation], our method [innovation] because [reason]"
- Avoid: "innovative," "novel," "groundbreaking" without specifics
APPROACH (research grant):
- Every method needs a rationale ("we chose X because Y")
- Every aim needs an alternative strategy ("if approach A fails, we will try B")
- Preliminary data is critical — shows feasibility, not just promise
TEAM:
- Match each team member's expertise to a specific aim
- Highlight track record: publications, previous grants, relevant experience
- For early-career PIs: explain how senior collaborators mitigate risk
Produce: criterion-by-criterion gap analysis + priority improvements.
Example
User: I'm applying for an NSF SBIR Phase I grant for an AI-powered drug interaction checker. $275K, 6 months. Help me write the Broader Impacts section.
Claude's Broader Impacts draft:
Broader Impacts
The proposed research addresses a significant and growing public health challenge. Adverse drug events (ADEs) caused by drug interactions affect an estimated 1.3 million Americans annually and account for over $5 billion in avoidable healthcare costs (AHRQ, 2023). Despite this burden, pharmacists and prescribers currently lack real-time, context-aware screening tools capable of evaluating complex polypharmacy scenarios — a gap that our technology is designed to address.
Healthcare access and equity. Our AI-powered drug interaction checker is designed for deployment in under-resourced clinical settings, including Federally Qualified Health Centers (FQHCs) and rural pharmacies where specialist consultation is unavailable. By providing guidance previously accessible only in well-resourced academic medical centres, this technology has the potential to reduce health disparities in medication safety across socioeconomic and geographic boundaries.
Workforce development. Phase I funding will support one full-time research engineer and one part-time clinical pharmacist collaborator. Both positions will gain cross-disciplinary experience at the intersection of machine learning and clinical pharmacology — skills in high demand as health systems increasingly adopt AI-assisted clinical tools.
Commercialisation and knowledge transfer. A successful Phase I will produce a validated prototype and a published accuracy benchmark dataset, made available as open data to the broader research community. This dataset, covering [X] drug pairs across [Y] clinical scenarios, will support future research in pharmacovigilance, drug repurposing, and clinical natural language processing.
Scientific infrastructure. The benchmark and evaluation methodology developed in this project will establish a replicable framework for assessing AI performance in clinical pharmacology — an area currently lacking standardised evaluation protocols. This contribution extends beyond our specific application to benefit the broader field of clinical AI validation.
Approximately 275 words — adjust to fit NSF's page limits.
