claude-skillz plugin
Nick Tune's plugin marketplace for Claude Code: checks before a task is handed in, code review and TDD personas
- 358 ★ on the repository
- Last commit
Extended access
Why this level
- Some plugins install hooks that fire on their own: review on session stop and a version check at start
- The session-optimizer and optimization-team plugins read your session transcripts
- fetching-circleci-logs calls the CircleCI API with your token
Install
- Works in
Manual install
/plugin marketplace add ntcoding/claude-skillzThen install the plugins you need: /plugin install task-check@claude-skillz. Available are task-check, automatic-code-review, full-codebase-review, track-and-improve, collaboration-modes, learn-from-prs, challenge-that, architect-refine-critique, session-optimizer, optimization-team, claude-code-updates, fetching-circleci-logs and the development-skills bundle.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Some plugins install hooks that fire on their own: review on session stop and a version check at start
- The session-optimizer and optimization-team plugins read your session transcripts
- fetching-circleci-logs calls the CircleCI API with your token
This is third-party code. Review the repository files before installing.
What it does
claude-skillz is a Claude Code plugin marketplace with Nick Tune's skills and system prompts. You add it with /plugin marketplace add ntcoding/claude-skillz and then install the plugins you want one by one with /plugin install name@claude-skillz. The list has eleven plugins plus a separate development-skills bundle. Several plugins act as checks. task-check looks for unfinished work, bugs and quality problems before a task is declared done. automatic-code-review runs a semantic code review when the session stops, using configurable rules, and full-codebase-review applies the same rules to the whole codebase. learn-from-prs analyzes feedback from CodeRabbit, SonarQube and reviewers and suggests config changes so such problems are caught sooner. Others set the way of working. collaboration-modes switches the style: step by step, deep thinking or pair programming. challenge-that judges a proposal from five adversarial positions, and architect-refine-critique runs a design through an architect, a refiner and a critique and produces an ADR. session-optimizer and optimization-team analyze session transcripts, and track-and-improve captures mistakes with a five-whys analysis. Separately there are skills and twelve personas started by the cl console launcher: Super TDD Developer among them. Skills attach to prompts by @ references.
Who it is for. For Claude Code developers who want ready-made completion checks, review on stop and personas for TDD and design.
Good fit when
- You want to switch the agent between step-by-step work, deep thinking and pair programming
- After every session you want an automatic semantic review of the changes against project rules
- You write code in a strict red-green cycle and want the Super TDD Developer persona with its eleven rules
- You want a proposal checked from five sides before implementation, for example before choosing an architecture
- CodeRabbit and SonarQube comments keep repeating and you want to turn them into config rules
Not a fit when
- You do not use Claude Code: the plugins and hooks target it, and for OpenCode there is only a generator that builds agents from the same personas
- You want one small skill: the marketplace holds many plugins and extra hooks load the session
- You need tools for a specific domain such as payments or databases: the set is general and about the development process
Example request
Run task-check on what you did in this session and list the loose ends before I accept the workLimitations
- The plugins install in Claude Code only. For OpenCode a generate-opencode-agents.py script builds agents from the same personas.
- The cl launcher that switches persona and model is set up by hand through an alias to claude-launcher.py, and interactive selection needs fzf.
- The repository states no license, so check the terms of use with the author.
- Some plugins install hooks that fire without your command.
- The repository was last updated in March 2026.
How to disable. Run /plugin, open the installed list and disable or remove the claude-skillz plugins you no longer want, one at a time. To drop the whole marketplace run /plugin marketplace remove claude-skillz. If you added the cl alias to your shell, delete it too.
FAQ
How do I connect claude-skillz to Claude Code?
Run /plugin marketplace add ntcoding/claude-skillz. Then open /plugin, choose Browse and install plugins and the claude-skillz marketplace, or install directly with /plugin install plugin-name@claude-skillz.
What does automatic-code-review do?
It runs a semantic code review when the session stops. On first install the plugin creates a default rule set, and it works with any language.
How does challenge-that work?
The /challenge-that command judges your proposal from five adversarial positions, including Skeptic, Pragmatist and Edge Case Hunter. You get objections before implementation starts.
What are cl and the personas?
cl is a console launcher: it starts Claude Code with a chosen system prompt and model, for example cl tdd opus. The twelve ready personas include Super TDD Developer, Strategic Architect, PRD Expert and Technical Investigator. You can add your own through frontmatter.
Can I install just one plugin?
Yes. Each plugin installs separately with /plugin install name@claude-skillz and leaves the rest untouched. The development-skills bundle collects general skills without hooks and commands.
Discussion
No questions or reviews yet.
Related
PluginsA marketplace of Claude Code plugins for engineering workflows: feature planning, tests, review, git automation and a code execution runtime
A Claude Code plugin marketplace by Frad LEE: Git and GitFlow, GitHub, refactoring, memory, research, image generation and interface work
A marketplace of 23 plugins for Claude Code: git and pull requests, code review, TDD, specs and delegation to other models
ClaudeKit Skills
ClaudeKit - Agent Skills
A plugin marketplace for Claude Code with a set of thirty-plus skills: frontend, backend, devops, databases, documents, debugging and MCP work
Claude Code Skills and Agents
Claude Code Skills & Agents
A Claude Code plugin bundling skills and sub-agents: deep research, image generation, GitHub work, code review and model switching
A plugin and skill marketplace for Claude Code: Cloudflare, frontend, design assets, writing, releases and integrations
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| claude-skillz (this entry) | no | 358 | |
| Superpowers | no | 307,845 | a strict development process from idea to merge, one plugin instead of a set of separate ones |
| project-starter | no | 1,420 | a universal plugin with seven agents and dozens of commands in one bundle |
README in short
The skills catalog has four groups: research, communication, design and process. The marketplace connects with one command, and the author lists plugins and personas separately.