Speclint – Score and fix your specs before AI agents build from them

Summary: Speclint evaluates specifications on five dimensions and scores them from 0 to 100, automatically rewriting sections that score below 70. It integrates via CLI, GitHub Action, MCP server, or REST API, ensuring specs are clear before AI agents generate code.

What it does

Speclint analyzes specs for measurable outcomes, testable criteria, constraints, vague verbs, and verification steps, then scores and rewrites failing parts automatically to improve clarity.

Who it's for

It is designed for developers and teams using AI coding agents who need precise, actionable specifications to avoid misinterpretation.

Why it matters

It prevents AI agents from building incorrect features by ensuring specs accurately reflect intended requirements before code generation.