R
robertmnyborg/claude-harness
One Claude Code marketplace for my skills, hooks, and session tools.
更新 2026-11-18
在 GitHub 查看 ↗直链 · 无中转页
装完之后怎么算装好了
- 装完该看到什么 — 命令正常结束时退出码为 0,终端里没有 "command not found" 或 "permission denied"。
- 一行自检 — 按该仓库 README 里给出的最小用法跑一次;能按 README 的描述输出就是通了。
- 答不出来怎么办 — 这个项目没有写排错说明,去仓库的 Issues 搜 install。
安装命令取自该仓库;以上判据只描述可观察到的结果,不替代项目自己的文档。
- ★ 数取值
- 0 ★ · robertmnyborg/claude-harness · 取自 GitHub 搜索 API · (2026-11-19 20:00:48 核对)
- 来源明细
- 从该仓库的 GitHub 页面直接读取 · 条目类型:开源项目
- 收录与活跃
- 收录于 2026-11-19 · 最近见到 2026-11-20 · 共见到 2 次 · 被 2 个来源列出
- 评分构成
- 综合分 203 · 2 个来源 · repo
- 安装
- 安装命令见上方「安装与配置」。
常见问题(答案都来自我们的记录)
它适合谁 / 用来做什么?
我们把它归在这些标签下: Agent 技能。标签来自条目自身,不是我们强加的分类。
谁维护它、还在更新吗?
仓库
robertmnyborg/claude-harness;我们收录于 2026-11-19;最近见到 2026-11-20。这里没有"官方认证"或"质量评级"——我们只记录来源与时间。README 原文(英文,未翻译) · 在 GitHub 读完整 README →
# claude-harness
以下摘录自该仓库的 README(片段)
Claude Code skills, hooks, and session tools from Robert Nyborg. This repo is the catalog: one marketplace that installs any of them.
/plugin marketplace add robertmnyborg/claude-harness
/plugin install skill-eval@claude-harness
Codex, Cursor, and other agents that read SKILL.md can use the skills CLI on any skill repo below: npx skills add robertmnyborg/ .
## The bar
Each repo starts from a failure I counted: a wrong answer an agent shipped, or a cost I measured. The tables below name the number where one exists.
Deterministic work runs as a script. The model handles judgment only.
Hooks and gates ship with tests you can run on your machine before you trust them.
I run skill-eval to decide what stays. A skill that scores worse loaded than unloaded gets cut.
## Install with the marketplace
| Plugin | What it does | Skills |
| skill-eval | Runs a skill with and without, judges both arms, reports the pass-rate delta. A public skill scored 77% loaded vs 97% without it. | skill-eval, retro |
| analyst-rigor-skills | Stops an agent from fabricating causal claims, root causes, quotes, personas, and strategies | causal-claim, metric-diagnosis, trustworthy-insights, job-card, strategy-check |
| everyday-agent-skills | Writes steps someone else can follow, judges a tool before you adopt it, decodes a confusing message, reads Reddit and X read-only | procedure-writing-rules, tool-fit, explain-message, social-research |
| ux-evidence | Makes the agent name the user, the pain, and the kill metric before it writes UI code | ux-evidence |
| product-thinking | Walks a non-engineer from an idea to a build-ready plan, one question at a time | product-thinking |
| self-review-evaluator | Rates a self-review against your leveling framework and plans the gap to the next level | self-review-evaluator |
| principal-reps | Timed interview reps that train the Principal-level PM answer pattern | principal-reps |
Install one with /plugin install @claude-harness.
## Install from the repo
These need scripts on your PATH or entries in settings.json, so each README walks you through it.
| Repo | What it does | The number behind it |
| claude-guardrails | Hooks that enforce rules a CLAUDE.md file can't | Chained sleep calls fell from 138 a week to 1 after a 10-line hook |
/plugin marketplace add robertmnyborg/claude-harness
/plugin install skill-eval@claude-harness
Codex, Cursor, and other agents that read SKILL.md can use the skills CLI on any skill repo below: npx skills add robertmnyborg/ .
## The bar
Each repo starts from a failure I counted: a wrong answer an agent shipped, or a cost I measured. The tables below name the number where one exists.
Deterministic work runs as a script. The model handles judgment only.
Hooks and gates ship with tests you can run on your machine before you trust them.
I run skill-eval to decide what stays. A skill that scores worse loaded than unloaded gets cut.
## Install with the marketplace
| Plugin | What it does | Skills |
| skill-eval | Runs a skill with and without, judges both arms, reports the pass-rate delta. A public skill scored 77% loaded vs 97% without it. | skill-eval, retro |
| analyst-rigor-skills | Stops an agent from fabricating causal claims, root causes, quotes, personas, and strategies | causal-claim, metric-diagnosis, trustworthy-insights, job-card, strategy-check |
| everyday-agent-skills | Writes steps someone else can follow, judges a tool before you adopt it, decodes a confusing message, reads Reddit and X read-only | procedure-writing-rules, tool-fit, explain-message, social-research |
| ux-evidence | Makes the agent name the user, the pain, and the kill metric before it writes UI code | ux-evidence |
| product-thinking | Walks a non-engineer from an idea to a build-ready plan, one question at a time | product-thinking |
| self-review-evaluator | Rates a self-review against your leveling framework and plans the gap to the next level | self-review-evaluator |
| principal-reps | Timed interview reps that train the Principal-level PM answer pattern | principal-reps |
Install one with /plugin install @claude-harness.
## Install from the repo
These need scripts on your PATH or entries in settings.json, so each README walks you through it.
| Repo | What it does | The number behind it |
| claude-guardrails | Hooks that enforce rules a CLAUDE.md file can't | Chained sleep calls fell from 138 a week to 1 after a 10-line hook |
收录徽章(可选)
想把收录标记放进 README 就复制这行,可选,不影响收录。
[](https://anclum.com/zh/s/robertmnyborg-claude-harness/)
- 来源
- github · https://github.com/robertmnyborg/claude-harness
- 按领域浏览
- Agent 技能
- 收录来源
- 2 个独立来源
- 收录于
- 今天
更多条目
所属领域与标签
每个标签都能点进去看同类的全部条目。数字是当前库里的真实条数。