improve-threejs

Audit and fix Three.js and React Three Fiber apps for frame-loop performance, GPU memory leaks, scene-graph correctness, and visual defects like z-fighting, shadow acne, wrong color space, and broken resize handling. Uses React Doctor as the scanning engine plus a visual rubric checked against rendered output. Use when the user asks to improve, audit, scan, or clean up a Three.js, R3F, react-three-fiber, drei, or WebGL app, or types `/improve-threejs`.

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

88%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, lean, and actionable audit workflow with explicit validation feedback loops. The main gap is actionability depth (template URLs and named-fix patterns rather than copy-paste recipes) and some inline content that could be split into a reference file.

Suggestions

Replace the template recipe URL with one or two concrete resolved example URLs so Claude can fetch a real recipe without guessing the plugin/rule segments.

Move the visual-audit rubric table and the code-level cause lists into a references/ file (e.g. visual-rubric.md) and link to it, keeping SKILL.md as a tighter overview.

For the highest-impact fixes (allocation inside useFrame, missing disposal), include a short copy-paste code snippet showing the before/after pattern.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence, with no padding explaining what Three.js, useFrame, or R3F are; every section (hot-path map, triage lists, visual rubric table, cause lists) earns its place.	5 / 5
Actionability	Provides concrete commands (npx react-doctor@latest --verbose, --scope changed), per-finding fix patterns, and a fail-condition rubric table, but the recipe URL is a template (<plugin>/<rule>.md) and fixes are named patterns rather than copy-paste code, leaving minor gaps below the fully-executable top anchor.	4 / 5
Workflow Clarity	Six clearly sequenced steps (Recon, Scan, Triage, Visual audit, Fix, Validate) with explicit validation checkpoints in Step 6 (re-run --scope changed, confirm no regression, re-check failed rows, watch memory profile) and a validate-fix-retry feedback loop, matching the top anchor.	5 / 5
Progressive Disclosure	Well-organized with clean section headers and only a one-level external recipe URL reference, but the dense visual-rubric table and dual cause lists are content that could plausibly live in a separate reference file, and the body exceeds 50 lines so the simple-skill 5 exception does not cleanly apply.	4 / 5
	Total	18 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A highly specific, well-triggered description that clearly answers both what the skill does and when to use it. It is distinguishable from generic React audit skills through its explicit Three.js/R3F and visual-defect focus.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions (audit and fix for frame-loop performance, GPU memory leaks, scene-graph correctness) and enumerates specific defect types (z-fighting, shadow acne, wrong color space, broken resize handling), giving comprehensive coverage rather than the minor gaps of a 4.	5 / 5
Completeness	Explicitly states what it does (audit and fix apps for enumerated issues, using React Doctor plus a visual rubric) and when to use it ('Use when the user asks to improve, audit, scan, or clean up... or types /improve-threejs'), with concrete trigger phrases matching the top anchor.	5 / 5
Trigger Term Quality	Covers natural terms with synonyms (Three.js, R3F, react-three-fiber, drei, WebGL) plus verb triggers (improve, audit, scan, clean up) and a slash-command trigger (/improve-threejs), exceeding the 'good coverage' bar of a 4.	5 / 5
Distinctiveness Conflict Risk	Occupies a clear Three.js/R3F performance-and-visual-audit niche with distinct 3D-specific triggers; conflict risk with other skills is minimal despite the React Doctor reference.	5 / 5
	Total	20 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: millionco/react-doctor
Path: skills/improve-threejs/SKILL.md
Commit: 83466a8

Reviewed: 5 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.