Anthropic Grade Optimizer

Audits and optimizes Claude-directing artifacts (CLAUDE.md, SKILL.md, subagent, hook, MCP config, system or user prompt, workflow, api_config) against 189 cited Anthropic rules across 11 dimensions, calibrated per model (Opus 4.7, Sonnet 4.6, Opus 4.6, Haiku 4.5). Every finding ships with verbatim quote and source URL; authorial voice is preserved.

20 stars 8 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add l0z4n0-a1/skill-anthropic-grade-optimizer --skill anthropic-grade-optimizer

Manual install

git clone --depth 1 https://github.com/l0z4n0-a1/skill-anthropic-grade-optimizer.git /tmp/skill-anthropic-grade-optimizer
mkdir -p ~/.claude/skills
cp -r /tmp/skill-anthropic-grade-optimizer/. ~/.claude/skills/anthropic-grade-optimizer

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Anthropic-Grade Optimizer

Audits any Claude-directing artifact against the official Anthropic doctrine,
calibrates findings by target model, and proposes surgical optimizations that
preserve authorial voice. Every finding cites a verbatim source URL; the skill
ships with cited rules only.

## Contents

- Hierarchy of authority
- Honest scope
- Quick start (and worked examples)
- The Three Laws
- Artifact types (9)
- Target-model modulation
- The 11 dimensions
- Severity → triage
- Emphasis conflict (D9, type-aware)
- Detection methods
- Operating modes
- Adaptive modes (auto-trigger)
- Output format (concise default)
- Workflow (10 steps)
- Scope discipline (positive framing)
- Edge cases
- Self-audit: Open Questions
- References (10 files)
- Assets (canonical snippets)
- Scripts (6 entry points)
- Validation (eval suite + strange-loop)

## Hierarchy of authority

Anthropic doctrine is the sole source of scoring rules. Optional interpretive
lenses inform *how* to reason about findings (what to preserve, when to defer,
how to phrase recommendations) — they stay outside the rubric. On collision,
Anthropic always wins.

## Honest scope

189 unique cited rules across 11 dimensions, each with a verbatim quote and
source URL. Read `references/rules-anthropic.yaml` § `meta.total_unique_rules`
for the canonical count — every other counter in the skill must read from
there to prevent drift. Rules with deterministic detection (regex,

Read the full skill on GitHub

Details

Skill name: anthropic-grade-optimizer
Repository: l0z4n0-a1/skill-anthropic-grade-optimizer
License: MIT
Repo updated: 3 May 2026

FAQ

What does the Anthropic Grade Optimizer skill do?

How do I install the Anthropic Grade Optimizer skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add l0z4n0-a1/skill-anthropic-grade-optimizer --skill anthropic-grade-optimizer. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Anthropic Grade Optimizer popular?

It lives in l0z4n0-a1/skill-anthropic-grade-optimizer, which has 20 GitHub stars and 8 forks.

Related skills

Chanlun Engine Skill

adsorgcn/chanlun-engine-skill

用缠论客观分析A股股票。当用户提到缠论、分型、笔、线段、中枢、背驰、买卖点、K线结构，或想知道某只股票该买该卖该持有还是等待时使用。确定性脚本计算缠论结构并给出四态操作判定(买入候选/持有/卖出减仓/等待观察)与失效价位，AI 只负责解读转述，不心算结构。免费数据源开箱即用，无需注册。

AI, LLMs & Agents

Alphaloop

realnaka/alphaloop

人机投研协作的操作系统——交易员出方向，Agent 出证据，决策权归人。它是一个编排器：把交易员抽象出的「逻辑框架」默认降级为待验证假设，依次调度 claim-verification（验真）、stock-data-fetch（取实时行情）、strategic-materials（产业链框架）、openorder（记忆/回写）、trade-journal（反馈闭环）、agent-tool-escalation（工具纪律），最后交还核验矩阵 + 个股清单 + 风险提示给交易员决策。当用户给出投资逻辑/趋势判断/受益方向、转发研报或小作文让你验证、问某公司/行业/标的、讨论估值或卡脖子或产业链、要选股或对比、要记一笔交易时，使用本 skill 串起整套流程。即使用户没点名"工作流"，只要进入投研语境就应当按这套分工与铁律执行。

AI, LLMs & Agents

Proactive Agent

halthelobster/proactive-agent

Transform AI agents from task-followers into proactive partners that anticipate needs and continuously improve. Now with WAL Protocol, Working Buffer for context survival, Compaction Recovery, and battle-tested security patterns. Part of the Hal Stack 🦞

AI, LLMs & Agents

Product Swipefile

nothingbutcici/product-swipefile

对一个明确命名的产品或产品集做深度产品调研（网站、App、SaaS、AI 工具、开源项目或创业公司），仅当用户用任意语言明确要求「调研 / 分析某个具名产品」时使用。

AI, LLMs & Agents

Project Blueprint

shuguang1994/project-blueprint

为新项目快速建立完整 AI 编程规范体系（AGENTS.md、文档目录、CI/CD、Git规范、测试制度）。自主发现引擎：扫描项目→分类文件→推断技术栈，覆盖 7 语言 15 框架 95 个组件条目，未知栈三层递进联网回退。Establish AI coding conventions for new projects — autonomous discovery engine with heuristic dep classification, 7 languages 15 frameworks 95 component entries, web search fallback for unknowns.

AI, LLMs & Agents

Oh My Skills

AIjunja/oh-my-skills

Use when the user wants to install, launch, inspect, inventory, create, edit, archive, deduplicate, or manage local Codex, Claude, or Agent Skills with the oh-my-skills GUI or CLI.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.