text-audit-pack

Binds the external text-audit skill to live-spec pack lints, paths, and reader rules.

1|Updated Jul 4, 2026
One-click install
npx skills add https://github.com/happysasha18/live-spec --skill text-audit-pack-happysasha18
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: text-audit-pack
Source: https://github.com/happysasha18/live-spec/tree/main/skills/text-audit-pack
Command: npx skills add https://github.com/happysasha18/live-spec --skill text-audit-pack-happysasha18

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When the external text-audit skill runs inside a live-spec project, it needs pack-specific context it no longer carries itself: which mechanical lint scripts to run, where reading records live, and what counts as a cheap reader under this pack's comprehension gate. This Skill supplies those bindings so an audit run inside the pack behaves consistently with the pack's own rules. ## Core Features & Use Cases - Pack lint bindings: Declares the six guardrail and lint scripts (vocabulary, weak words, requirement shape, style, register, one-name checks) that text-audit's Step 1 runs via .text-audit/lints.json before falling back to bundled grep. - Pack path registry: Points reading records to docs/language-reads/ and links the comprehension gate in spec-author to the audit loop. - Cheap reader definition: Defines that a cheap reader inside this pack is a reader with none of the pack's skills or rules loaded, so reading rounds count toward the comprehension gate only when run clean. - Use Case: A maintainer runs text-audit on a spec inside a live-spec project; loading this page beside the external skill's SKILL.md ensures the pack's own lints execute and the reading record lands in the right place. ## Quick Start Load the external text-audit skill's SKILL.md first, then load this page beside it, and run a text audit on a spec document inside the live-spec project.

Frequently Asked Questions about text-audit-pack

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run text-audit inside a live-spec project?▼

Load the external text-audit skill's own SKILL.md first, then load this pack-bindings page beside it. The audit run then executes the six pack lint scripts named in .text-audit/lints.json before falling back to the skill's bundled grep.

What lint scripts does the live-spec pack declare for text audits?▼

The pack declares six scripts: check-vocabulary.py, check-weak-words.py, check-requirement-shape.py, spec-style-lint.py, preshow-register-lint.py, and check-one-name.py. They cover undefined terms, weak relational words, missing requirements genre, style and register, and duplicate naming.

Can text-audit-pack run an audit on its own?▼

No. This page only binds the external text-audit skill to the pack; it runs no cold read and fixes no line itself. Without the external skill's SKILL.md loaded beside it, its lint table names commands but starts none.

What counts as a cheap reader in a live-spec text audit?▼

A cheap reader is a reader with none of the pack's own skills or rules loaded, including live-spec-base and any working skill. A session still carrying pack skills does not count as cheap, and its reading rounds do not count toward the comprehension gate's two clean rounds.

Why do new language rules not reach text-audit's bundled rule sheet?▼

After the 2026-08-18 extraction, text-audit's reference pages ship as a frozen static snapshot in the external repository. Rules added to guardrails/language-rules.json still reach the pack's own documents but no longer reach the audit skill's bundled sheet.