Lists multiple specific concrete actions: formatting, review criteria, budget preparation, broader impacts, significance statements, innovation narratives, and compliance with submission requirements. Also names specific agencies (NSF, NIH, DOE, DARPA, Taiwan NSTC).

The 'what' is well-covered with specific actions and agencies, but there is no explicit 'Use when...' clause or equivalent trigger guidance. The when is only implied by the nature of the tasks described, which caps this at 2 per the rubric guidelines.

Includes strong natural keywords users would say: 'research proposals', 'NSF', 'NIH', 'DOE', 'DARPA', 'budget preparation', 'broader impacts', 'significance statements'. These are terms researchers naturally use when seeking grant writing help.

Highly distinctive niche targeting competitive research grant proposals for specific funding agencies. The combination of agency names (NSF, NIH, DOE, DARPA, Taiwan NSTC) and grant-specific terminology (broader impacts, review criteria) makes it very unlikely to conflict with other skills.

Extremely verbose at 500+ lines. Extensively explains concepts Claude already knows (what a PDF is, what active voice means, what a Gantt chart is, basic writing advice like 'use strong verbs'). Lists of common mistakes, writing principles, and persuasive argumentation strategies are generic knowledge that wastes token budget. The 'When to Use This Skill' section is entirely unnecessary.

Despite its length, the skill provides almost no executable, copy-paste-ready content. It's overwhelmingly descriptive and advisory rather than instructional. There are no concrete proposal templates with actual text, no example passages showing good vs. bad writing, and no specific formatting commands. The one code block is for generating schematics, not for grant writing itself. The content reads like a textbook chapter, not actionable instructions.

The 5-phase workflow for grant development is clearly sequenced with outputs listed for each phase, which is good. However, there are no validation checkpoints or feedback loops within the writing process itself (e.g., no 'verify compliance with page limits before proceeding' or 'check alignment between aims and budget'). The workflow is more of a project management timeline than an operational procedure with decision points.

References to external files (references/*.md, assets/*.md, scripts/) are well-signaled and one level deep, which is good. However, no bundle files were provided, so these references may be broken. The main SKILL.md itself is a monolithic wall of text that should have pushed most agency-specific details into the referenced files rather than duplicating extensive content inline.