What coaching checks

Coaching evaluates each field against the grant's actual requirements and flags issues by priority:

• High — likely to matter to a reviewer; address before submitting.
• Medium — worth strengthening.
• Recommended — a nice-to-have improvement, not a gap.

A field that already meets the bar gets no suggestions at all — coaching doesn't manufacture feedback just because a field exists.

Two different kinds of issue on AI-drafted fields

This is the part worth understanding well, because the two categories call for different responses:

• Missing — the AI didn't have enough information to answer this fully and is asking you to supply it.
• Unverified — the AI generated something plausible-sounding that it couldn't actually confirm from your organization's data — a number, a claim, a specific detail it inferred rather than sourced.

Treat these differently. A "missing" suggestion means: add the information. An "unverified" suggestion means: check it, correct it, or remove it — don't just build more content around it. The single most common mistake is treating an unverified claim like a missing one and elaborating around a number or fact that was never actually confirmed.

When you'll see coaching

Coaching appears automatically on AI-drafted fields as soon as your application is generated — you don't have to click into a field or start typing to get feedback on it. It also re-checks a field once you edit it, so if you rewrite a suggestion away, you'll know whether your new version holds up.

‍