Clarity, precision, structure — measured.
Six traits learned from the best technical writing. Sub-second scoring, no signup.
Content score a URL example
Paste a link to a doc, blog post, or tutorial
Analysis

What gets measured

Usable guidance
Whether the writing gives the reader something they can immediately do, apply, or run.
actionable steps vs passive description
Problem-anchored
Whether each concept is introduced in response to a visible problem or limitation.
motivated by real problems vs unmotivated exposition
Evidence-backed precision
Whether the writing uses specific numbers, real examples, and concrete scenarios.
specific, verifiable claims vs vague handwaving
Layered build-up
Whether concepts are introduced one at a time and composed step by step.
progressive disclosure vs complexity dump
Coherent arc
Whether sections connect into a narrative that builds toward a conclusion.
connected narrative vs disconnected sections
Concrete-first
Whether examples and working code appear before abstract explanations.
examples lead vs abstract-first

What do the scores mean? →

Write → score → improve

The ⊨ Claude Code plugin turns scoring into a feedback loop. Evaluate your draft, then let the agent iterate until traits converge.

1
/u22a8:evaluate
Score your current file against technical-writing traits. See exactly where clarity drops.
2
/u22a8:improve
The agent rewrites weak sections and re-scores until all traits cross the threshold you set.

Install from the plugin marketplace →

MCP server

Streamable HTTP — works with Claude Code, Cursor, Windsurf, or anything that speaks MCP.

# Claude Code $ claude mcp add --transport http u22a8 https://u22a8.ai/mcp

From your terminal

$ cat post.md | curl -s -d @- \ -H "Accept: text/plain" \ https://u22a8.ai/p/u22a8.technical-writing

Full API docs →

Scores are approximate — not a substitute for human judgment.
built by @onebit0fme