Strayguard review
Built to sniff out what you never told us.
Our AI Grant Writer drafts your application. Before you see it, it's sent to Strayguard, our second AI reviewer. Strayguard scores the draft against what the funder says it scores, and flags details it can't trace back to your answers. You confirm them or cut them.
Relentless by design. Even the best guard dog misses one now and then, so read every line.
Scout finds the grants that fit you. Our AI Grant Writer writes the application. Strayguard reviews it before you see it.
How it works
You tell us about your business.
Your answers become the record Strayguard checks against. Nothing else counts as proof.
Strayguard reads the draft.
It asks one question of every claim about your business: did you tell us this, or did the writer add it? It also scores the draft against the funder's published criteria.
You decide what stays.
Anything it flags is highlighted in your draft. Mark it right, or edit it. If the score is low, we try one rewrite and give you the stronger version.
The kind of thing Strayguard™ catches
- A customer story you never told us.
- A revenue number, headcount, or award you never gave.
- A booking system or waitlist you don't actually keep.
- “Four years in business” when it's really three.
- A promise you never made, like covering a cost from revenue.
- Budget lines and forecasts you never gave us go on a separate list for you to confirm.
A real flag, from a sample business
Michelle Johnson's med spa is a fictional sample business. The draft and this flag are real output from an earlier run of our system. The draft said:
“On a busy day, one of them is waiting while the other treats.”
Flagged: You told us you have one room and two injectors. You did not tell us one waits while the other works. Confirm it is true, or cut it.
Our secret sauce
Every draft is sent to Strayguard before it reaches you. When the funder lists its scoring criteria, Strayguard scores your draft against each one. If the score comes back low, we try one rewrite and give you the stronger version. What it flags is right there in your draft. How it decides is our secret, and it stays that way.
What Strayguard™ won't do
- Catch everything. It's strict, and it still misses things. Read your draft before you submit.
- Know when an answer is wrong. It checks the draft against your answers. If an answer is off, the draft will be too.
- Decide if you qualify. The funder decides that.
- Delete things for you. What it flags stays highlighted until you confirm or cut it.
- Always finish. Once in a while the review can't complete. When that happens, your draft page says so plainly, and you should read that draft with extra care.
- Pretend to be a person. Strayguard is AI. It runs on Claude, made by Anthropic, with our own review rules.
Questions
- Is Strayguard a person?
- No. Strayguard is a second AI, separate from the one that writes your draft. It runs on Claude by Anthropic. You review and submit the application yourself.
- Does it catch every invented detail?
- No, and we won't claim it does. That's why everything it flags comes to you, and why you should read your draft before you send it.
- What happens when it flags something?
- The detail is highlighted in your draft. Tap it to jump there, then mark it right or edit it. Anything the draft proposes that you never told us, like a budget line, goes on a separate 'confirm these' list.
- What if the review doesn't finish?
- Your draft page tells you, right at the top. Read that draft closely before you submit, and email hello@winagrant.com if you want help.
- Why don't you publish how it works?
- Because it's our secret sauce. We show you what it catches. The rules stay ours.
Strayguard comes with every application. $25 for one, or $97 for 10.
Find grants that fit you