Spec Score

A balanced spec produces balanced code.
An unbalanced spec produces creative fiction.

Install

# Add to Claude Code
npm i -g spec-score-mcp
claude mcp add spec-score -- npx spec-score-mcp

4 Axes

completeness
Can the LLM understand the full scope?
clarity
Is there only one way to interpret this?
constraints
Does the LLM know what NOT to build?
specificity
Are there concrete, testable details?

Verdicts

SHIP ITSpec is ready — the LLM knows what to build and what not to
ALMOSTOne axis needs a small fix before you start
DRAFTMultiple axes need work, but the structure is there
VAGUEWell-organized but too abstract to act on
UNBOUNDEDClear goal but no boundaries — the LLM will over-build
OVER-CONSTRAINEDLots of rules but unclear what the actual goal is
SKETCHStarting point — needs detail on most axes

Get it

npm v2.0.0
npx spec-score-mcp
GitHub
openpoem/spec-score-mcp — source, docs, slash commands
Smithery
MCP server registry listing
ChatGPT Custom GPT
Use Spec Score directly in ChatGPT

API

POST /api/score
Score a spec → JSON with balance, verdict, axes
POST /api/visualize
Score a spec → SVG radar chart
POST /api/compare
Compare two specs → side-by-side SVG