Hosho
/ Prompt Intelligence
Pricing
Docs
Research
Agent guide
Hosho Research
What actually makes a prompt work - and what the evidence says.
The messy problem of prompting
Why small wording changes swing outcomes - and why your coding agent can't fix it
July 2026
So what matters in a prompt?
Four factors, twelve sub-factors, and the research behind each
July 2026
Catch fragile prompts before you ship or switch models
Hosho reads prompt text for a score, line-anchored findings, and fixes - MCP, Prompt Editor, or Git Bot beside your evals
September 2026
Versioning answers what shipped. Review answers why it is fragile.
Why a versioned, eval-tested prompt can still ship fragile text - and the lighter read that catches it
September 2026