Guide · Agent skill
How to use the Ego-Lite Simplify agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The ego-lite-simplify skill is a set of written instructions that tells an AI coding agent how to review the ego-lite repository for simplification opportunities such as dead code, duplicated state and unnecessary compatibility layers. It is aimed at maintainers and contributors of that repository, and at teams who want a worked example of an evidence-driven cleanup checklist. After this guide you will know what the skill changes about agent behaviour, how to install it, how to phrase requests that trigger it, and what it needs from your machine.
Ego-Lite Simplifycitrolabs/ego-lite · MIT Download the skill Quick start
- 1Get the ZIP. Download the ego-lite-simplify ZIP from this page.
- 2Unzip into a skills folder. Extract it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder.
- 3Or upload in the Claude apps. Go to Settings → Capabilities → Skills and upload the skill there instead.
- 4Prepare the ego-lite checkout. Clone or open a local checkout of the ego-lite repository with git and ripgrep (rg) available.
- 5Ask for an audit. Describe the cleanup goal in plain language and ask for a report before any changes.
- 6Review the report. Check the cited caller evidence and decide which candidates to approve.
- 7Run the checks. Let the agent run the relevant npm test, npm run e2e or npm run validate:site-skills commands and git diff --check.
What is an agent skill, and what does this one make the agent do?
An agent skill is a folder containing a SKILL.md file with instructions that an AI agent loads when a request matches the skill's description. This skill's description says it applies when reviewing the codebase for dead code, duplicated state or APIs, speculative abstractions, unnecessary compatibility layers, hand-written infrastructure, excessive tests or documentation, or when the user asks to reduce code size or maintenance complexity without hiding behavior changes.
Once loaded, the skill directs the agent to read AGENTS.md and the design document for the affected behaviour, check git status so unrelated changes are preserved, and search for real callers with rg before judging any code. Candidates are classified as production, contract, non-production or ambiguous evidence. The agent is asked to report worthwhile candidates before any broad removal, and to separate behaviour-changing ideas that need a user decision from behaviour-preserving cleanups.
Install the skill from the ZIP
Download the ZIP for the skill from this page. It contains the skill files, which are agents/openai.yaml and SKILL.md. Unzip it so that the folder containing SKILL.md ends up in one of the following places.
For a personal install that applies across your projects, place it in ~/.claude/skills/. For a single project, place it in that project's .claude/skills/ folder. Alternatively, if you use the Claude apps, you can upload the skill under Settings → Capabilities → Skills instead of copying files.
- Personal: ~/.claude/skills/ego-lite-simplify/
- Project: .claude/skills/ego-lite-simplify/ inside your project
- Claude apps: Settings → Capabilities → Skills, then upload
How do you trigger the skill with a natural request?
You do not need a special command. Describe the goal in plain language and mention the ego-lite repository, so the request matches the skill's description. Requests about reducing code size or maintenance complexity without hiding behaviour changes are the clearest triggers.
Ask for an audit first if you want to decide what changes. The skill explicitly supports an audit mode, where the agent reports candidates and makes no broad removals, and an authorized implementation mode, where it makes the change.
- Audit: "Review the ego-lite repo for dead code and unnecessary compatibility layers. Report candidates with evidence, don't change anything yet."
- Implement: "In ego-lite, remove the unused export from the runtime I confirmed with you, using the test-first process in the simplify skill."
- Size and maintenance: "Reduce maintenance complexity in ego-lite without changing public behaviour."
What do you need before running it?
The skill assumes a local checkout of the ego-lite repository. The test commands it names are npm test, npm run e2e and npm run validate:site-skills, which run from the package/ego-browser/ directory and need Node.js and npm. It also uses git for status and diff checks, and ripgrep (rg) for searching callers. The skill runs scripts and does not use the network, according to its listing.
Because the skill refers to specific files such as AGENTS.md and package/ego-browser/skill-body/SKILL.md, it is only useful against that repository.
- Local checkout of the ego-lite repository
- Node.js and npm
- git
- ripgrep (rg)
Safety notes: scripts and what the agent may run
The skill is marked as running scripts. The verification steps it describes run the repository's npm test scripts and finish with git diff --check. Review the commands the agent proposes before approving them, and run the agent in a checkout where you can inspect the changes with git.
The instructions tell the agent not to edit generated build output as the source of a simplification, and not to remove a defensive path until its trust or lifecycle boundary is understood. These are guardrails in the instructions, not technical enforcement, so keep a human review step before merging any broad removal.
- The skill lists scripts as used and network as not used.
- Check git status before and git diff after each run.
- Treat the skill's rules as guidance to the agent, not as a sandbox.
Tips for getting useful results
Ask for a small, bounded area such as one module or one kind of candidate. The skill itself advises preferring a few well-proven deletions over a long list of guesses, so narrow requests tend to produce better evidence.
When a candidate is a behaviour or API decision, the skill expects it to be routed to the owning design document first. Have that document available so the agent can read it.
- Name the module or surface you care about.
- Ask for evidence (caller searches) in the report.
- Keep public API changes as a separate decision from cleanups.
What people use it for
- Finding exports, helpers or options in ego-lite that have no production caller, and confirming with rg before removal.
- Spotting duplicated state, such as two representations of the same session or page, and proposing which to keep.
- Reviewing whether hand-written utilities could be replaced by a Node.js built-in, with a net-size estimate.
- Producing an audit report that separates behaviour-preserving cleanups from changes needing a user decision.
- Running a test-first cleanup in ego-lite, updating characterization tests before removing obsolete code.
Frequently asked questions
How do I install the ego-lite-simplify skill?
Download the ZIP from this page and unzip it so the folder containing SKILL.md sits in ~/.claude/skills/ for personal use or in a project's .claude/skills/ folder. You can also upload it in the Claude apps under Settings → Capabilities → Skills.
What does ego-lite-simplify require?
You need a local checkout of the ego-lite repository, Node.js and npm for the test commands, git, and ripgrep (rg) for searches and status checks. The skill is specific to that repository and its file paths.
Is the ego-lite-simplify skill free, and what is its licence?
The item is listed under the MIT licence. The source provided does not state any price, so do not assume anything beyond the licence listed.
Does the skill use the network or run scripts?
The listing says it runs scripts and does not use the network. Its verification steps run npm test, npm run e2e, npm run validate:site-skills and git diff --check, so review those commands before approving them.
Can I use this skill on a project other than ego-lite?
Not as intended. The file names, paths and commands are specific to the ego-lite repository, so it is not a general cleanup tool. You can read it as an example of an evidence-driven refactoring checklist for your own repository.
Will the skill clean up my code automatically?
No. The skill asks the agent to prove each candidate with caller searches and to report before broad removal. Behaviour-changing ideas are meant to be decided first, so it is not designed for unreviewed automatic cleanup.
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
This skill suits maintainers and contributors of the ego-lite repository who want a disciplined, evidence-first cleanup pass, and teams looking for a model checklist. It is a poor fit if you work on any other project, or if you want quick automatic cleanup without review, since it deliberately asks for evidence and reports before removing anything.
Download the skill