Structure the review the editor actually wants

Structure the review the editor actually wants

A good review is a service to the field and it takes a Saturday. The manuscript is fine; deciding what is a major concern and what is taste is the part that eats the time.

Why the mainstream APIs fail here

Reviews get ignored for shape, not substance: no separation between fatal and cosmetic, no statement of what would fix each concern. A model helps you structure the criticism you already made, and it does not get squeamish about a paper on snake venom the way general chatbots do.

How MagmaRouter is used for this

Paste the manuscript and your notes; ask for concerns sorted into major, minor and comment, each with what would resolve it. Ask whether your tone would survive an author who disagrees. The manuscript is not retained.

A working loop

  1. Write your concerns raw first; the model sorts, it does not read your mind.
  2. Every major concern needs the fix stated, not just the flaw.
  3. Say what is good, specifically; vague praise reads as a form letter.

Prompts are not stored. A signed receipt of the hash comes back on every call, so you can prove later what MagmaRouter logged. Details are on the privacy page. Per-model prices are on /models and the pricing model is on /pricing.

Related