Skip to content

Guide · Agent skill

How to use the Skillify agent skill: turn workflows into skills

Updated · written with AI from the project's own documentation and checked against our catalogue data

Skillify is an agent skill from the Oh My Claude Code (OMC) project that turns a multi-step process your agent just worked out into a reusable skill draft. It is for developers and teams who keep re-solving the same project-specific problem and want the next session to start from a saved playbook. After this guide you will know what the skill does, how to install it, how to trigger it, and what to review before trusting the output.

Skillify: Turn Workflows into SkillsYeachan-Heo/oh-my-claudecode · MIT

Quick start

  1. 1Finish the workflow first. Let the agent complete the multi-step process in the session so the steps and results exist to be captured.
  2. 2Install the skill. Unzip the downloaded ZIP into ~/.claude/skills/ for all projects, or into .claude/skills/ for one project, or upload it under Settings → Capabilities → Skills in the Claude apps.
  3. 3Ask in plain language. Tell the agent the workflow is worth saving, for example by naming the task you just completed.
  4. 4Let the quality gate run. The agent checks that the workflow is repeatable, project-specific and effortful; generic how-tos should go to documentation instead.
  5. 5Review the draft. Open the generated file, check the steps, success criteria and pitfalls, and answer any listed open questions.
  6. 6Save it to a location. Keep it as a user-level file under omc-learned or a project file under .omc/skills/, and commit project skills so teammates can use them.

What does the Skillify skill make the agent do?

Skillify tells the agent to check whether a workflow from the current session is worth keeping, and if it is, to write it down as a skill file. The skill's stated goal is to capture a successful multi-step workflow as a concrete skill draft instead of rediscovering it later.

Before extracting anything, the agent applies a quality gate. The workflow should be something no one could quickly Google in five minutes, specific to this codebase, project or workflow, and something that took real debugging, design or operational effort to find. Generic snippets, boilerplate and library usage examples are meant to stay in normal documentation.

  • Identifies the repeatable task the session accomplished
  • Extracts inputs, ordered steps, success criteria, constraints and pitfalls, verification evidence, and a target location
  • Decides whether the result is a repo built-in skill, a user/project learned skill, or documentation only
  • Drafts a skill file that starts with YAML frontmatter, including name, description and triggers
  • Lists open questions that are still too fuzzy to encode safely

How do you install Skillify?

There are two routes described on this page. For Claude Code, download the Skillify ZIP from this page and unzip it into your user skills folder, ~/.claude/skills/, or into a project's .claude/skills/ folder if you want the skill limited to one repository.

If you use the Claude apps instead, upload the skill under Settings → Capabilities → Skills. The package contains a single file, SKILL.md, so after unzipping you should see that file inside the skill's folder.

  • Option A (Claude Code, all projects): unzip into ~/.claude/skills/
  • Option B (Claude Code, one project): unzip into .claude/skills/ inside the project
  • Option C (Claude apps): upload via Settings → Capabilities → Skills

How do you trigger Skillify with a natural request?

The skill is meant for moments right after a session has worked something out. You do not need a special command. Describe what you want in plain language once the agent has finished the process, and the skill's description tells the agent when it applies: turning a repeatable workflow from the current session into a reusable OMC skill draft.

Be specific about the workflow you want captured. Naming the task helps the agent identify which steps belong in the draft.

  • "That deploy-and-verify process we just went through is worth saving as a skill."
  • "Turn the steps we used to fix the flaky integration tests into a reusable skill draft."
  • "Skillify the database migration workflow from this session."

What does a generated skill draft look like?

A learned skill is a complete file that begins with YAML frontmatter. The minimum frontmatter the skill asks for is a name, a one-line description, and a list of triggers. The body then covers the steps, success criteria and pitfalls.

The skill specifies where learned skills are written: a flat file under ${CLAUDE_CONFIG_DIR:-~/.claude}/skills/omc-learned/<skill-name>.md for user-level skills, or .omc/skills/<skill-name>.md for project-level skills. Uncommitted project skills stay local to your worktree until they are committed or copied to a user-level directory.

  • Frontmatter: name, description, triggers
  • Body: clear triggers, ordered steps, success criteria, pitfalls
  • Storage: omc-learned is kept as a directory name for compatibility, not as the name you invoke
--- name: <skill-name> description: <one-line description> triggers: - <trigger-1> - <trigger-2> ---

What should you review before keeping a draft?

Treat the output as a draft. The skill itself notes that a draft may encode incorrect assumptions from one session. Check that each step is actually correct, that the success criteria can be verified, and that the pitfalls reflect real experience rather than guesses.

The skill also asks the agent to point out anything still too fuzzy to encode safely, and to note unresolved branching decisions before drafting. Answer those open questions before you rely on the file.

  • Confirm each ordered step works as written
  • Check that success criteria are concrete and testable
  • Remove or correct pitfalls the session did not really confirm
  • Keep the file only if the workflow is repeatable

What does Skillify need, and is it safe to run?

Skillify is part of the Oh My Claude Code (OMC) framework, so that framework is a requirement. It runs in Claude Code or another agent runtime that can read SKILL.md skill folders. The package contains only SKILL.md.

According to this listing, the skill runs no scripts and uses no network access. It works by guiding the agent to write a markdown file with frontmatter to a folder on your machine. Since the output is a file you can open and edit, you stay in control of what gets saved and reused.

  • Requires: OMC framework, Claude Code or a compatible agent runtime
  • Runs scripts: no
  • Uses network: no
  • Files: SKILL.md only

What people use it for

  • Saving a project-specific debugging playbook after a hard-won fix
  • Recording the ordered steps for a recurring deployment or migration process
  • Building a shared team library of project conventions for OMC
  • Capturing verification steps so a later session knows how to confirm success

Frequently asked questions

Is Skillify free to use?

This listing shows the skill under the MIT licence. It is part of the Oh My Claude Code project, and this page does not list any pricing.

What do I need to run the Skillify skill?

You need the Oh My Claude Code (OMC) framework, which Skillify is part of, plus Claude Code or another agent runtime that can read SKILL.md skill folders.

Where do I install the Skillify skill in Claude Code?

Unzip the download into ~/.claude/skills/ to make it available everywhere, or into .claude/skills/ inside a project to limit it to that project.

How do I install Skillify in the Claude apps?

Upload the skill in the Claude apps under Settings → Capabilities → Skills, using the downloaded ZIP.

Does Skillify run scripts or access the internet?

No. This listing says the skill runs no scripts and uses no network access. It only guides the agent to write a skill file.

Can I use Skillify for generic how-tos or library examples?

The skill is designed not to do that. It says generic snippets and library usage examples belong in ordinary documentation, and it only targets repeatable, project-specific workflows.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Skillify suits developers and teams who repeatedly solve the same project-specific problems with an agent and want to save those workflows as reviewable drafts. Skip it for one-off fixes, generic how-tos, or anyone who expects output that is production-ready without human review.

Download the skill