Twenty prompts that hold up across nonprofit grant-writing tasks. Each is written as a fill-in template you copy into your AI tool and adapt. All prompts assume you’ll review and substantively edit the output before it goes to a funder. Last reviewed: May 25, 2026.

Ground Rules for Using These

• Do not paste participant names, SSNs, dates of birth, or other identifiers into any AI tool that is not enterprise-grade and approved for that data class.
• Treat every output as a first draft. Substantively edit. Verify any numerical, regulatory, or factual claim against an authoritative source before submission.
• Disclose AI assistance per your organization’s Staff Handbook disclosure standard — most grant narratives sit at Footer or Explicit tier.
• If a funder’s instructions prohibit AI-assisted drafting, do not use these prompts on that proposal.

Phase 1: Funder Research

Prompt 1 — Funder fit assessment

I am a grant writer at [Organization Name]. Our mission is [one sentence]. Our current programs serve [population] in [geography]. Based on the funder information I’m pasting below, give me: (1) a one-paragraph fit assessment — plausible, marginal, or poor fit — with the specific reasons; (2) the three strongest angles I could lead with in a proposal; (3) anything in their public materials that suggests a poor fit I should not push against. Do not invent details about the funder that aren’t in what I pasted.

Funder information: [paste public materials — mission, recent grants, RFP language, IRS Form 990 program services description].

Prompt 2 — Public-record summary

Summarize the following funder materials into a one-page internal briefing covering: stated mission and priorities; geographic focus; typical grant range if visible; application process and deadlines if stated; past grantees the funder has highlighted; anything in the materials that suggests application timing or relationship-building requirements. Use only what’s in the text below. If something is unclear or missing, list it under “Open questions” rather than guessing.

Materials: [paste].

Prompt 3 — Question list for a funder conversation

I have a 20-minute introductory call with a program officer at [funder name] next week. Based on the funder summary I’m pasting, draft eight to ten questions I could ask that would (a) give the program officer useful information about our work without rambling, (b) help me learn what specifically about our work would interest them, and (c) avoid asking anything that the funder’s public materials already answer. Mark which questions are essential and which are optional.

Funder summary: [paste].

Phase 2: Needs Statement

Prompt 4 — Needs statement outline

Draft an outline (not full prose) for a needs statement of approximately [word count]. The statement should establish: (1) the specific problem we’re addressing; (2) who is affected, in our service area; (3) what authoritative data confirms the problem; (4) why current responses are insufficient; (5) what we’re proposing. Use the facts I’m pasting below; do not introduce statistics or quotes I haven’t given you. Note any places where adding a recent local data point would strengthen the section.

Facts: [paste — program data, community-needs assessment, public data citations].

Prompt 5 — Local data integration

Below is a draft needs statement and three local data points. Suggest how to integrate the data points into the narrative — specifically, where each one fits, what framing makes it land, and what context the funder would need to read it correctly. Do not invent new data. If a data point seems out of place, say so.

Draft: [paste].

Data points: [paste with sources].

Prompt 6 — Tone calibration

This needs statement was written to be honest. Read it and tell me where it might come across to a foundation reader as: (a) defensive, (b) under-confident, (c) over-claiming, or (d) emotionally manipulative. Suggest specific edits, not rewrites. Preserve the voice and the substance.

Statement: [paste].

Phase 3: Theory of Change

Prompt 7 — Theory of change from program description

Below is our program description. Reconstruct an honest theory of change as: Inputs → Activities → Outputs → Short-term Outcomes → Long-term Outcomes → Impact. Use only what’s in the program description. Mark any link in the chain that the description does not actually support; do not paper over weak links with plausible-sounding language.

Program description: [paste].

Prompt 8 — Assumptions audit

Below is a theory of change. List every assumption it makes that connects one stage to the next — the things that have to be true for the chain to hold. Rate each as well-supported, plausible, or untested. Suggest one piece of evidence we could add for each plausible-or-weaker assumption.

Theory of change: [paste].

Phase 4: Evaluation Plan

Prompt 9 — Indicators from outcomes

Below are our proposed outcomes. For each, suggest one to three indicators that would credibly demonstrate the outcome, with: the data source, the collection frequency, who is responsible, and the threshold of meaningful change. Avoid indicators we can’t actually collect.

Outcomes: [paste].

Capacity context: We have [team size and data systems available]. We do not have access to [common limitations].

Prompt 10 — Evaluation methodology paragraph

Draft a one-paragraph evaluation methodology for a [size] proposal that proposes [program]. The evaluation will be [internal / external / mixed]. We will use [data sources]. The reporting cadence is [quarterly / annual]. Be specific about what we will measure, how, and by when — not abstract about “continuous improvement.”

Prompt 11 — Honest limitations

For the evaluation plan below, list the three most material limitations a reasonable evaluator would name — selection bias, attribution, attrition, measurement reliability, etc. For each, suggest a brief one-sentence acknowledgment we could include in the proposal that would not undermine the proposal but would not paper over the limitation.

Plan: [paste].

Phase 5: Budget Narrative

Prompt 12 — Budget narrative from a budget table

Below is our project budget. Draft a budget narrative that, line by line, explains: what the cost is, why it’s necessary for the funded activity, how the amount was determined, and (if shared with other funding) how it’s allocated. Do not change any numbers. Flag any line where the rationale isn’t clear from the budget alone so I can add context.

Budget: [paste].

Prompt 13 — Uniform Guidance check

For a federally funded proposal under 2 CFR Part 200, review the following budget for cost categories that often draw monitor scrutiny: indirect costs vs. direct treatment, allocation methodology for shared costs, salary and wage support tied to time-and-effort documentation, equipment thresholds, travel, contract vs. consultant treatment. Identify lines that need additional support in the narrative. This is not legal or audit advice — only a checklist of things to address.

Budget: [paste].

Prompt 14 — In-kind contributions

Below is a list of in-kind contributions we plan to claim. For each, suggest how to describe and value it in the narrative: source of the valuation, basis for the rate, documentation we will need to maintain. Note any item whose valuation is hard to defend and would be safer to drop.

In-kind list: [paste].

Phase 6: Letter of Inquiry

Prompt 15 — Two-page LOI from a full proposal

Compress the attached full proposal into a two-page Letter of Inquiry covering: brief organizational background, the problem, our proposed work, why we are positioned to do it, requested amount and use, and a one-sentence ask. Cut adverbs and unnecessary qualifiers. Do not invent new claims; remove anything the LOI cannot support.

Proposal: [paste].

Prompt 16 — Opening paragraph variants

Draft three different opening paragraphs for this LOI. Variant A: anchored in a single specific participant story (composite, no identifiers). Variant B: anchored in a single piece of authoritative data. Variant C: anchored in our most distinctive program design choice. For each, write only the first 60 to 80 words. Note which variant is the strongest fit for [funder type] and why.

LOI draft: [paste].

Phase 7: Revisions and Submission

Prompt 17 — Funder-instruction compliance check

Below are two documents: (1) the funder’s RFP or proposal instructions, and (2) our current draft. Check the draft for compliance with the funder’s instructions — required sections, word limits, page limits, attachments, formatting, signature requirements, due date, submission method. Produce a checklist with each requirement and whether the draft meets it. Do not improve the prose; only check compliance.

Instructions: [paste].

Draft: [paste].

Prompt 18 — Reviewer red-flag scan

Read this draft as a foundation program officer who reviews 40 proposals a month. Flag the three places where you would mark it down, the three places where it loses you, and the one thing you would most want to ask the applicant. Do not rewrite — just diagnose.

Draft: [paste].

Prompt 19 — Clarity pass

Edit the following passage for clarity. Preserve the voice, do not change any factual content, do not soften any of the specifics. Cut adverbs that don’t add information, replace abstract verbs with concrete ones where the meaning is clear, and break up sentences longer than 35 words unless the length is doing real work.

Passage: [paste].

Prompt 20 — Post-decision debrief

We received a [funded / declined / partially funded] response from [funder] for our [proposal]. The funder’s feedback was: [paste, or note “no feedback given”]. Draft a one-page internal debrief covering: what worked, what did not, what we would do differently if we approached this funder again, and whether to reapply. Be specific. Do not invent feedback the funder didn’t give.

A Note on Honesty

Every prompt in this library includes an instruction to use only what you provided and to flag unknowns rather than fabricate. That is the single most important pattern. AI tools will confidently produce plausible-sounding statistics, organizational details, and quotes if you don’t constrain them. Constrain them.

---

Pair with Writing Stronger AI Grant Applications and the Staff Handbook disclosure standard.