Hikaruegashira Agent Skills
hikaruegashira-agent-skills is a Claude Code plugin marketplace for the Build phase that bundles three session-tested agent skills for structure, review, and sandbox-aware agent workflows.
by HikaruEgashira · github.com/HikaruEgashira/agent-skills
Install recycled agent skills from real AI sessions to structure projects, enforce sandbox allowlists, and run architect-style reviews in Claude Code.
Add it to Claude Code
Install the plugin in Claude Code. One command, paste-ready.
/plugin install hikaruegashira-agent-skills@HikaruEgashira/agent-skillsBuilt to be called by your agent
Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.
Retrieve this entry with skillselion.get_details("plugin:HikaruEgashira/agent-skills") and the paste-ready config with skillselion.get_install_config("plugin:HikaruEgashira/agent-skills").
What it does
hikaruegashira-agent-skills is a small Claude Code plugin marketplace that packages three skills the author refined across real AI coding sessions. Solo builders who live in Claude Code and want opinionated agent behavior—not generic prompts—can add the marketplace and pull in plugins focused on project structure, architectural thinking, and review loops. The surfaced keywords emphasize agents, commands, sandbox boundaries, and domain allowlists, which suggests skills aimed at keeping autonomous work inside declared scopes while still moving fast. That combination fits indie builders shipping agents or automation-heavy products where one mis-clicked network call is costly. Complexity is intermediate: you need Claude Code’s plugin workflow and comfort tuning sandbox rules. It is not a full IDE replacement or a hosted MCP server; it is a curated skill set you register and invoke like any other marketplace plugin.
Highlights
- 3 Claude Code plugins distilled from the author’s AI working sessions
- Agent-oriented workflows: architect, project management, and review-style skills
- Sandbox and allowlist-oriented controls (alloweddomains, sandboxblocked) for safer agent execution
- Command and domain-management patterns for repeatable agent behavior
- Community marketplace entry—register once to browse all bundled skills
Why builders use it
Generic agent defaults make it hard to reuse the rituals that actually worked in your past sessions—structure, safe domains, and review—without rebuilding prompts every project.
After registering the marketplace you can install all three plugins and apply consistent agent commands, allowlists, and review patterns across repos without rewriting skills from scratch.
At a glance
- Type - Plugin in AI Agents.
- Adoption - 0 installs, 1 stars, 0 votes.
FAQ
Who is hikaruegashira-agent-skills for?
It is for Claude Code users who want a small set of agent skills—project structure, review, and sandbox allowlisting—captured from the author’s real sessions.
When should I use hikaruegashira-agent-skills?
Use it when you are in the build phase configuring agent tooling and need repeatable commands and domain rules instead of ad-hoc prompts each sprint.
How do I add hikaruegashira-agent-skills to my agent?
Add the HikaruEgashira/agent-skills marketplace in Claude Code, then install the individual plugins you want from the bundle of three.
Comments
Share how you use hikaruegashira-agent-skills, gotchas, or tips for other indie builders.
No comments yet - be the first to share how you use it.