Voice Validator — Data & AI agent for Claude Code
Runs the post-generation checks on writer-style output: a fact-preservation diff (every verified number/identifier must survive styling unchanged), an AI-tell lint (banned words, false-antithesis, em.
How to install Voice Validator
Installs to ~/.claude/agents/solanabr-writer-style-skill-voice-validator.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/solanabr/writer-style-skill/HEAD/agents/voice-validator.md -o ~/.claude/agents/solanabr-writer-style-skill-voice-validator.md Restart Claude Code, or start a new session, for it to be picked up.
What Voice Validator does
name: voice-validator description: "Runs the post-generation checks on writer-style output: a fact-preservation diff (every verified number/identifier must survive styling unchanged), an AI-tell lint (banned words, false-antithesis, em-dash overuse, uniform cadence), and a repetition audit across a batch of lessons. Reports pass/fail with the specific offending lines; does NOT rewrite — it gates.
Use when: a draft (or a batch) has been produced and needs the Pass-C gate before shipping; ver
Alternatives in Data & AI
- Read Research For Nyquist Validation Data — cat "$phase_dir"/-RESEARCH.md 2>/dev/null node "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" roadmap get-pha 38.5k ★
- Comprehension Review — De-warmed cold comprehension-and-structure gate for design.md and track files: reads only the document, never 432 ★
- Identifier Validator — Use this agent practively whenever new identifiers (PMIDs, DOIs, database IDs, ontology term IDs) are introduc 163 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Platform SQL Reviewer
Critique the SQL/data pipeline behind a headline number (contamination, aggregation bias, query performance) a
Prose Review
Reviews prose for AI patterns, banned phrases, voice drift, and structural issues. Dispatched by the copydesk
AI Evaluation Engineer
Measure AI quality honestly with datasets, metrics, and judges that survive scrutiny. Invoke for data-ai-team
Design Guardian
Accountable design decider and applier. Reviews real UI evidence, rejects token drift and generic AI styling,
Example: Incremental Data Pipeline Pattern
from datetime import datetime, timedelta @dag(schedule="@daily", catchup=False) def incremental_sales_pipeline
Review Arbiter
Authoritative LLM-as-a-Judge meta-agent that adjudicates multi-agent plan and code reviews whenever two or mor
Related Skills
Xhs Compliance Skill
Claude Code Agent Skill for XHS (小红书) content compliance — avoid banned words and traffic limits
Docblock Rewrite
Convert PHPDoc and JSDoc blocks into one-line plain-English // comments. Scripted bulk pass over a directory o
Advanced Code Review Verify
Advanced Code Review Phase 4: Verification - fact-check findings against codebase, remove false positives