Does this skill help with broken links?

Yes, it automatically runs Python QA scripts to detect broken internal links, invalid anchors, and markdown files missing from the book's summary.

How does the skill define review standards?

The skill uses a master standards file at `.claude/book-review/standards.md` and individual policy files in the same directory to guide parallel specialist agents.

How are deferred issues handled?

Issues marked as 'Defer' during triage are automatically formatted and appended to `book/FIXME.md` for future resolution, avoiding duplicate entries.

What happens if multiple agents find the same issue?

The skill includes a synthesis step that organizes findings by location and merges overlapping concerns into a unified list of must-fix issues and suggestions.

Can I review specific parts of a book instead of the whole thing?

Yes, you can specify file paths, specific line ranges (e.g., file.md:10-20), or even conceptual topics which the skill will locate using grep.

Technical Book Reviewer

Name: Technical Book Reviewer
Author: tachyon-zcash

bytachyon-zcash

•

Learning & Documentation

Performs automated, multi-agent editorial reviews of technical documentation using parallel specialist agents and custom style policies.

The Book Review skill is a sophisticated documentation auditor designed specifically for technical books like the Ragu proof-carrying data framework. It orchestrates a fleet of parallel AI agents to evaluate content against specific style guides and standards, while simultaneously running Python-based QA scripts for link validation and formatting checks. The skill manages the entire editorial lifecycle—from interpreting complex review scopes (like git diffs or conceptual topics) to validating proposed fixes and managing a deferred issue backlog in a central FIXME file.

Key Features

01Change Validation: Pre-validates all suggested rewrites against project standards to ensure fixes don't introduce new policy violations.

0233 GitHub stars

03Intelligent Scope Resolution: Interprets git diffs, specific file paths, line ranges, or conceptual topics to define review boundaries.

04Automated QA Integration: Runs internal scripts to detect broken links, dead pages, and line-width violations in Markdown files.

05Parallel Multi-Agent Review: Orchestrates multiple specialist agents to check documentation against distinct policy files simultaneously.

06Integrated Triage Workflow: Allows users to instantly apply fixes, skip findings, or defer issues to a structured tracking file.

Use Cases

01Automating the maintenance of large mdBook projects by identifying broken internal anchors and unlinked pages.

02Performing cross-chapter reviews of complex technical concepts to ensure narrative coherence and eliminate redundancies.

03Auditing uncommitted documentation changes for consistency with existing technical terminology and style guidelines.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tachyon-zcash/ragu book-review

For use in Claude.ai and ChatGPT

Download Skill