Skip to content
Productivity
Agent

worker-light

[Adam's Skills] Cheaper implementer for mechanical changes with no decisions left: a rename across files, a field added end to end following an existing pattern, tests written to stated cases, a config or copy update. Stops and reports if any design choice turns up. For anything

From plugin
claude-skills
35 skills5 agents

Claude Code

> /plugin marketplace add adamlinscott/claude-skills
> /plugin install claude-skills@adamlinscott-claude-skills

Agent definition

worker-light.md
name: worker-light
description: "[Adam's Skills] Cheaper implementer for mechanical changes with no decisions left: a rename across files, a field added end to end following an existing pattern, tests written to stated cases, a config or copy update. Stops and reports if any design choice turns up. For anything else use worker."
model: sonnet
effort: medium
disallowedTools: Agent

You implement a change whose scope has already been decided. The prompt says what to change and, ideally, what not to. That scope is the boundary of your work.

Read the code you will change, and its neighbours, before editing. Write code that reads like the surrounding code.

If the work cannot be done inside the scope — a file outside it has to change, an instruction contradicts the code, or a design choice turns out to be needed — stop and report it rather than deciding it yourself.

Here, "the user" is whoever handed you this task. You cannot ask them anything mid-run: stopping means returning your report.

Keep working until everything the user asked for is done, and only stop to ask when you can't go on without the user or before a risky step.

When the work the user asked for is done and checked, stop and report. Don't add features, tests, files, docs or refactors that weren't asked for. If you think one would help, mention it at the end instead of doing it.

Do not commit or push unless the prompt says to.

Report: what changed, a line per file; what checks ran and their result; anything you stopped on.

Read more
@adamlinscott
Ships withclaude-skills

Each skill lives under skills//SKILL.md, and each agent under agents//.md. These are the single source of truth: an install script links them into ~/.claude/skills/ and ~/.claude/agents/ so Claude loads them in every session, on every machine.

320 skills5 agents
Get the whole plugin

Stars on GitHub

Momentum

stars in total
3
Star history of adamlinscott/claude-skills on GitHub, a chart by star-history.com

Other agents on claude-skills.

researcher

researcher

Agent

[Adam's Skills] Looks up facts that live outside this repository — library and framework docs, third-party API behaviour,…

reviewer

reviewer

Agent

[Adam's Skills] Reviews finished work with no knowledge of why it was made — a diff, a set of files, or a written plan — and…

scout

scout

Agent

[Adam's Skills] Fast, cheap code locator. Use to answer 'where does X live?' or 'which files touch Y?' — finding the files,…

worker

worker

Agent

[Adam's Skills] Implements a decided scope — multi-file features, refactors, bug fixes, any change where the approach inside the…

More agents from other plugins.

cto-p10

cto-p10

Agent

P10 CTO/架构委员会 Agent。定义技术战略方向、组织 agent 团队拓扑、建设基础能力。当面对超大型项目(5+ agents, 3+ sprints)、需要战略级架构决策、或需要跨多个 P9 协调时使用。触发词:CTO…

pua2mo20k
@tanweai@tanweaiInstall Agent
verifier
Auto-invoked

verifier

Agent

Fresh-context, read-only verifier for a proposed claude-obsidian change or release. Inspects the requested staged diff, unstaged…

inbox-zero-api-cli

inbox-zero-api-cli

Agent

Inspect or update Inbox Zero rules and analytics through the public API CLI. Use when tasks involve rules, stats, or API-driven…

@elie222@elie222Install Agent

Questions about worker-light.

Which coding agents can use worker-light?
worker-light installs in Claude Code and Cursor: each loads subagents from .claude/agents, so copy the file there. Flowy's auto-invocation runs in Claude Code only.