Teach agents to blueprint fundable grant narratives before layout: a falsifiable hypothesis, explicit gaps, independent aims, reviewer-scannable hierarchy, and honest feasibility—without inventing data or drowning reviewers in prose.
This skill produces grant packages as
documents. Follow the
k2f skill for authoring: unpack or init a workspace, edit the semantic tree, run pack/verify, then export. K2F keeps agents on
meaning (section roles, aims, evidence)—not pixel coordinates—while the engine compiles layout, typography, and theme into consistent output across PDF, DOCX, and other formats.
Before you build, install the core skill:
npx skills add suzheng/k2f --skill k2f
.
Single blueprint for semantic content and aesthetic intent. Structure it with labeled blocks the k2f skill can map to pages later.
-
Mechanism lock — Agency, solicitation ID, mechanism, page limits, required sections, scoring criteria (NIH: Significance, Innovation, Approach, Investigators, Environment; NSF: Intellectual Merit, Broader Impacts). Drop optional material that exceeds limits.
-
Four-question audit — Answer in prose using user-supplied evidence only:
- Central hypothesis: testable, falsifiable (not “we will study X”).
- Why now: gap, urgency, new tools or data.
- Innovation: conceptual, methodological, application, or novel outcomes—each with specifics.
- Feasibility: team, timeline, resources, aim independence, pitfalls with Plan B.
-
Specific Aims architecture — One page, ~400–550 words. Narrative arc: hook → gap → long-term goal → hypothesis → 2–3 verb-led aims (testable, largely independent) → impact close. Per aim: 2–3 sentences, expected outcome, no fatal dependency on prior aim success.
-
Strategy outline — Significance (problem, state of field, gap, cost of not knowing, if-successful impact). Innovation as scannable bullets (“This project is innovative because…”). Approach per aim: rationale → strategy/controls/stats → expected outcomes → pitfalls/alternatives → milestones. Assign word budgets from page limits (~450 words per full strategy page); note where figures replace paragraphs (preliminary data, workflow, timeline).
-
Design intent — Document tone, hierarchy, density, palette for the engine:
- Tone: formal, evidence-led; a non-expert primary reviewer grasps significance on page 1.
- Hierarchy: Aims page = strongest contrast and largest headings; strategy H1s conclusion-driven (“Gap in X limits Y”) not generic labels.
- Density: bullet-led Innovation; Approach subheads per aim; no paragraphs over ~6 lines.
- Palette: calm academic (navy/slate body, one accent for aim numbers and key metrics); WCAG AA; no decorative clutter on the Aims page.