将任何会话转化为可重复使用的技能。 Skillify 会监控你在对话中的操作——你使用的工具、你所做的修正、你遵循的工作流程——并将其捕获为便携式 agentskills.io SKILL.md 文件。

支持Claude Code、Cursor、GitHub Copilot、Gemini CLI、VS Code以及支持 agentskills.io 标准的其他30+代理平台。

它的作用

在你做了可重复操作的会话结束时，运行/skillify，它会：

1. 重建发生了什么 — 回顾对话历史、git修改和项目背景
2. 面试你 — 4轮结构化轮次，捕捉意图、步骤、成功标准和边缘案例
3. 生成 SKILL.md——一个可移植的技能文件，遵循 agentskills.io 开放标准
4. 保存 — 保存到你的项目（.claude/skills/）或个人目录（~/.claude/skills/）

生成的技能在下一次游戏中立即可以作为斩击指令使用。

安装

克劳德·科德

然后重启Claude Code。该技能将以/skillify形式开放。

手动安装（任何代理平台）

将skills/skillify/SKILL.md复制到你的代理人技能发现目录中：

/skillify v1.0.0

Turn any session into a reusable skill. Skillify watches what you did in a conversation — the tools you used, the corrections you made, the workflow you followed — and captures it as a portable agentskills.io SKILL.md file.

Works with Claude Code, Cursor, GitHub Copilot, Gemini CLI, VS Code, and 30+ other agent platforms that support the agentskills.io standard.

What It Does

At the end of a session where you did something repeatable, run /skillify and it will:

1. Reconstruct what happened — reviews conversation history, git changes, and project context
2. Interview you — 4 structured rounds to capture intent, steps, success criteria, and edge cases
3. Generate a SKILL.md — a portable skill file following the agentskills.io open standard
4. Save it — to your project (.claude/skills/) or personal directory (~/.claude/skills/)

The generated skill is immediately usable as a slash command in your next session.

Installation

Claude Code

# Copy the skill to your personal skills directory
mkdir -p ~/.claude/skills/skillify
curl -sL https://raw.githubusercontent.com/kk-r/skillify-skill/main/skills/skillify/SKILL.md \
  -o ~/.claude/skills/skillify/SKILL.md

Then restart Claude Code. The skill will be available as /skillify.

Manual Installation (Any Agent Platform)

Copy skills/skillify/SKILL.md into your agent’s skill discovery directory:

Install Script

bash <(curl -sL https://raw.githubusercontent.com/kk-r/skillify-skill/main/scripts/install.sh)

Usage

Basic — Capture Current Session

After completing a workflow you want to repeat:

/skillify

Skillify reviews your session, asks clarifying questions, and generates a SKILL.md.

With Description — Seed the Interview

/skillify deployment workflow for staging

Providing a description helps Skillify focus on the right parts of your session.

How It Works

Phase 0: Context Reconstruction

Since SKILL.md-based skills don’t have direct access to session memory (that’s a bundled-skill-only capability), Skillify reconstructs context from three sources:

1. Conversation history — reviews all messages to identify goals, steps, corrections, and tool usage
2. Git artifacts — git diff --stat and git log to see what changed
3. Project detection — auto-detects package.json, Makefile, Cargo.toml, go.mod, etc. so generated skills reference the right tooling

Phase 1: Structured Interview

Four rounds via interactive prompts (not plain text):