Guide · Agent skill
How to use the CE Skill Work agent skill (ce-skill-work)
Updated · written with AI from the project's own documentation and checked against our catalogue data
The ce-skill-work skill is an agent skill from the EveryInc compound-engineering-plugin repository. It teaches an AI agent to work on skill files under skills/**, following the repository's written authoring standard. This guide explains what the skill does, how to install it, how to trigger it with plain requests, and what it needs before it is useful.
CE Skill WorkEveryInc/compound-engineering-plugin · MIT Download the skill Quick start
- 1Download the ZIP. Download the ce-skill-work ZIP from this page.
- 2Unzip the skill. Unzip the file so the ce-skill-work folder contains SKILL.md and its references.
- 3Choose a location. Place the folder in ~/.claude/skills/ for personal use or .claude/skills/ inside a project.
- 4Or upload in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the skill.
- 5Confirm the standard is present. Check that docs/solutions/skill-design/portable-agent-skill-authoring.md exists in your repository.
- 6Make a natural request. Ask for a skill create, edit, or review in plain words that describe the file under skills/**.
What is an agent skill, and what does ce-skill-work do?
An agent skill is a folder containing a SKILL.md file and optional reference files. The agent reads the skill when a request matches its description, then follows the instructions in it.
ce-skill-work applies the repository's skill-authoring standard as a procedure. It covers creating a skill, editing one, reviewing a skill change, and acting on review feedback from a human or a bot about a skill. The skill's stated goal is that skill files state their conditions rather than list many cases, carry only lines that change behavior, and are validated in proportion to their risk.
- Creating a new skill, starting from an outcome spine written as prose
- Editing an existing skill while checking the block's goal before changing it
- Reviewing a skill change against the standard
- Responding to review comments about a skill
- Scope: files under skills/**. It is not meant for src/, tests/, or scripts/ code.
Install ce-skill-work from the ZIP download
Download the ZIP for ce-skill-work from this page and unzip it. Place the resulting skill folder in one of two locations. Use ~/.claude/skills/ to make it available across your projects, or use the .claude/skills/ folder inside a single project to limit it to that project.
If you use the Claude apps instead of a folder, you can upload the skill in Settings → Capabilities → Skills.
- Personal install: ~/.claude/skills/
- Project install: .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills, then upload the skill
Trigger the skill with a natural request
You do not need a special command. Describe the job in plain words that match the skill's description, and the agent should load it. The description covers creating a skill, editing one, reviewing a skill change, and acting on review feedback about one.
Requests about application code, tests, or build scripts fall outside the skill's scope, so the agent should not rely on it for those.
- Create a new skill for handling release notes in skills/ and follow the authoring standard.
- Edit skills/ce-plan/SKILL.md so the rule about done conditions is stated as a condition, not a list of cases.
- Review this pull request that changes skills/**/SKILL.md and tell me which lines do not change behavior.
- Address the reviewer comments on my skill change and explain which ones you applied.
What the skill needs before it works well
The skill depends on a referenced standard document, docs/solutions/skill-design/portable-agent-skill-authoring.md, and on project instruction files that supplement it. The skill reads these from the repository. It does not include them, so it is designed for work inside the compound-engineering repository or a project that carries the same files.
The skill runs no scripts and makes no network calls, according to its listing.
- The repository's portable-agent-skill-authoring.md standard document
- Project instruction files that the skill says supplement the standard
- Runs scripts: no
- Uses network: no
Safety notes on scripts and network access
Because the skill runs no scripts and uses no network, it does not execute code or send data out on its own. Its effects come from the file edits the agent makes in your repository and any validation the agent chooses to run.
Review the diff the agent produces before you commit, as you would for any change to skill files.
- No bundled scripts are run by this skill
- No network use is declared
- Check edits to skills/** before committing
Tips for getting good results
Give the agent the skill file you care about and say whether you want a create, edit, or review. The skill has separate reference files for each of those jobs, so a clear mode helps it choose the right one.
Keep the standard document in the repository. Without it, the skill has nothing to apply.
- Name the exact file under skills/** you are changing
- State whether you want creation, editing, review, or a response to feedback
- Make sure the standard document exists in your checkout
What people use it for
- Creating a new skill under skills/ with an outcome spine written first
- Editing an existing SKILL.md so its rules state conditions rather than enumerated cases
- Reviewing a pull request that changes skill reference files
- Acting on bot or human review comments about a skill change
Frequently asked questions
Is ce-skill-work free to use?
The listing states the skill is licensed under MIT. Check the license file in the source repository for the full terms.
What do I need to install ce-skill-work?
Download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it in the Claude apps under Settings → Capabilities → Skills. The skill also expects the repository's portable-agent-skill-authoring.md standard document to be present.
Does ce-skill-work run scripts or use the network?
No. The listing says it runs no scripts and uses no network.
Can I use ce-skill-work for writing application code?
No. The skill is scoped to skill files under skills/** and excludes src/, tests/, and scripts/ code. For those, use other guidance.
Why does the skill say it depends on a standard document I do not have?
The skill applies a written authoring standard and points to it rather than restating it. Without the document and the project instruction files, the skill lacks the rules it applies, so it is best used inside the compound-engineering repository or a project that includes those files.
How is this different from a general skill-writing guide?
ce-skill-work is a procedure for one repository's standard. It walks the agent through creating, editing, reviewing and responding to feedback on skills, rather than offering general advice on writing any skill format.
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
Use ce-skill-work if you maintain or contribute to the compound-engineering plugin, or if you want a worked example of a strict skill-writing procedure. Pick something else if you need help with application code, tests, or build scripts, or if you want a skill format that is not tied to this repository's standard.
Download the skill