Skillify

Skillify

You are capturing this session's repeatable process as a reusable skill.

Your Task

Step 1: Analyze the Conversation

Before asking any questions, analyze the conversation to identify:

• What repeatable process was performed
• What the inputs/parameters were
• The distinct steps (in order)
• The success artifacts/criteria for each step (e.g., not just "writing code," but "an open PR with CI fully passing")
• Where the user corrected or steered you
• What tools and permissions were needed
• What the goals and success artifacts were

Step 2: Interview the User

Use AskUserQuestion for ALL questions. Never ask questions via plain text. For each round, iterate as needed until the user is happy. The user always has a freeform "Other" option — do NOT add your own "Needs tweaking" option.

Round 1: High-level confirmation

• Suggest a name and description for the skill based on your analysis. Ask the user to confirm or rename.
• Suggest high-level goal(s) and specific success criteria for the skill.

Round 2: More details

• Present the high-level steps as a numbered list. Tell the user you'll dig into detail in the next round.
• If the skill needs arguments, suggest them based on what you observed.
• If unclear, ask if this skill should run inline (in the current conversation) or forked (as a sub-agent). Forked is better for self-contained tasks; inline is better when the user wants to steer mid-process.
• Ask where to save:
  • This repo (.claude/skills/<name>/SKILL.md) — for project-specific workflows
  • Personal (~/.claude/skills/<name>/SKILL.md) — follows you across all repos

Round 3: Breaking down each step For each major step, if not obvious, ask:

• What does this step produce that later steps need? (data, artifacts, IDs)
• What proves this step succeeded?
• Should the user confirm before proceeding? (especially for irreversible actions)
• Are any steps independent and could run in parallel?
• How should the skill be executed? (direct, Task agent, agent team)
• What are the hard constraints or preferences?

You may do multiple rounds here, especially with 3+ steps. Pay attention to user corrections during the session.

Round 4: Final questions

• Confirm when this skill should be auto-invoked. Suggest trigger phrases.
• Ask about any gotchas.

Don't over-ask for simple processes.

Step 3: Write the SKILL.md

Create the skill directory and file at the user-chosen location.

Use this format:

---
name: {{skill-name}}
description: {{one-line description}}
allowed-tools:
  - {{list of tool permission patterns}}
when_to_use: {{when Claude should auto-invoke, with trigger phrases and examples}}
argument-hint: "{{hint showing argument placeholders}}"
arguments:
  - arg1
  - arg2
context: {{inline or fork — omit for inline}}
---

# {{Skill Title}}

Description of skill.

## Inputs

- `$arg_name`: Description of this input

## Goal

Clearly stated goal. Define artifacts or completion criteria.

## Steps

### 1. Step Name

What to do. Be specific and actionable.

**Success criteria**: What proves this step is done.

**Artifacts**: (optional) Data produced for later steps.

**Human checkpoint**: (optional) When to pause for user confirmation.

**Rules**: (optional) Hard constraints.

...

Frontmatter rules:

• allowed-tools: Minimum permissions (use patterns like Bash(gh:*) not just Bash)
• context: fork: Only for self-contained skills that don't need mid-process user input
• when_to_use is CRITICAL — start with "Use when..." and include trigger phrases
• arguments and argument-hint: Only include if the skill takes parameters

Step structure tips:

• Concurrent steps use sub-numbers: 3a, 3b
• User actions get [human] in the title
• Keep simple skills simple — 2-step skills don't need every annotation

Step 4: Confirm and Save

Before writing, output the complete SKILL.md as a code block for review. Ask for confirmation using AskUserQuestion: "Does this look good to save?"

After writing, tell the user:

• Where the skill was saved
• How to invoke: /{{skill-name}} [arguments]
• That they can edit the SKILL.md directly to refine it