conversation-workflow-packager

name	conversation-workflow-packager
description	Create a reusable SKILL.md from chat history. Use when converting a repeated process into a project skill with decision points, quality checks, and example prompts.
argument-hint	What recurring workflow should this skill package?
user-invocable	true
disable-model-invocation	false

Conversation Workflow Packager

What This Skill Produces

This skill produces a complete SKILL.md draft that turns a repeated chat workflow into a reusable, project-scoped skill.

When to Use

• You notice the same step-by-step method repeated across requests.
• You want to preserve review or implementation checklists as a slash-invocable workflow.
• You need a first version quickly, then refine ambiguous decision points.

Procedure

1. Review the current chat and extract:

• Ordered steps that are repeated
• Decision points with branching logic
• Quality gates that define completion

2. If the workflow is not clear, infer a minimal default flow and mark assumptions.
3. Create a workspace skill at .github/skills/<skill-name>/SKILL.md.
4. Write frontmatter with:

• name that matches folder name
• A keyword-rich description with "use when" triggers
• Optional argument-hint if slash use benefits from input guidance

5. Write the body with:

• Outcome and scope
• Trigger conditions
• Step-by-step process
• Decision rules
• Completion checks
• Example prompts

6. Identify weak or ambiguous areas and ask focused follow-up questions.
7. Incorporate answers and finalize with a short "how to use" section.

Decision Rules

• If workflow appears project-specific, store under .github/skills/.
• If workflow is personal and cross-project, recommend a personal skills directory.
• If the process is single-shot and parameterized, suggest a prompt file instead of a skill.
• If the process must be always-on, suggest instructions instead of a skill.

Completion Checks

• name matches folder name exactly.
• Description contains concrete discovery keywords.
• Steps are executable without hidden knowledge.
• Decision points are explicit.
• At least 3 example prompts are included.

Example Prompts

• /conversation-workflow-packager Turn our bug triage routine into a reusable skill.
• /conversation-workflow-packager Package our release checklist into a team skill.
• /conversation-workflow-packager Convert our code review method into a slash command workflow.

Follow-on Customizations

• Add a paired prompt in .github/prompts/ for one-off skill drafting.
• Add instructions in .github/instructions/ for naming and frontmatter conventions.
• Add a custom agent if this becomes a multi-file customization pipeline.