Figure Critic — Data & AI agent for Claude Code
Reviews a generated or existing figure (data plot or diagram) against figure-quality rules before it goes in the paper — checks it actually has a single clear message, is referenced and interpreted in.
How to install Figure Critic
Installs to ~/.claude/agents/hverma99-phd-skills-figure-critic.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/hverma99/phd-skills/HEAD/agents/figure-critic.md -o ~/.claude/agents/hverma99-phd-skills-figure-critic.md Restart Claude Code, or start a new session, for it to be picked up.
What Figure Critic does
name: figure-critic description: Reviews a generated or existing figure (data plot or diagram) against figure-quality rules before it goes in the paper — checks it actually has a single clear message, is referenced and interpreted in text, and has a self-sufficient caption. tools: Read, Grep model: sonnet
You are a figure critic for a chemical/process systems engineering research paper or thesis chapter. You review one figure at a time — a data plot or a conceptual/architecture diagram
Alternatives in Data & AI
- Kiro Steering — agent-style v0.4.2 — literature-backed technical-prose writing rules for AI agents 650 ★
- Review Code Style — A code review agent that evaluates code quality, readability, naming conventions, DRY violations, dead code, a 134 ★
- Prose Review — Reviews prose for AI patterns, banned phrases, voice drift, and structural issues 70 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Mermaid Flow Visualizer
Use this agent when the user explicitly requests a Mermaid diagram to visualize code architecture, data flow,
Paper Density
Review academic papers for information density. Finds redundancy, repeated data, duplicate presentations (tabl
Claim Grader
Use when a contributor wants a soft-gate review of a commit message, PR body, sprint section, or journal entry
Logo Critic
Blind, independent scorer for logo candidates, especially AI-generated ones. Use after any generation round or
Read Only Explorer
Maps how code actually works — call paths, call sites, data flow, existing patterns, ownership boundaries, and
Migration Safety Reviewer
Use this agent after every dotnet ef migrations add and before any migration lands in a commit. Reviews the ge
Related Skills
Subtext
A Claude Skill that tells you what a message is actually saying, and shows its work. Evidence-tiered subtext a
Claude Study Skills
Claude Code skills for actually learning things — Socratic tutoring, lecture breakdowns, depth-calibrated expl
Doc Figure
제안서·공문서·사업기획안 삽도를 스타일 고정 그림으로 만드는 Claude Code 스킬 — 한글이 깨지지 않는 코드 렌더, 템플릿 11종·프리셋 3벌(공공/공문서 흑백/기획안), API 키·설치 불