davidsimoes

Claude Skills Shared — Development skill for Claude Code

Development community

Curated Claude Code skills from David's setup — MVP for SGA team, seed for broader open-source release.

How to install Claude Skills Shared

This entry records only its repository, not the path inside it, so there is no exact command to give. Open davidsimoes/claude-skills-shared and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Claude Skills Shared does

Curated Claude Code skills from David's setup — MVP for SGA team, seed for broader open-source release.

Alternatives in Development

Knowledge Work Plugins — Open source repository of plugins primarily intended for knowledge workers to use in Claude Cowork 9734 125 18 10.5k ★
Llamacoder — Open source Claude Artifacts – built with Llama 3.1 405B 7.1k ★
Microsandbox — Open-source self-hosted MicroVM sandboxes with sub-200ms startup, hardware-level isolation via libkrun 5k ★

README

claude-skills-shared

Eleven skills for [Claude Code](https://docs.claude.com/en/docs/claude-code) — small, focused tools that change how an agent works in your terminal. Each is a single `SKILL.md` (plus supporting scripts/templates where needed) that Claude Code auto-loads when its description matches what you're asking for.

![respond-html demo artifact](./docs/respond-html-demo.png)

*Above: `respond-html` rendering a structured reply with reactable units. Inline ✅/💬/❌ buttons on every section/callout, live counts in the sticky toolbar, one-click copy as markdown. One of the 11 skills shipped here.*

What's a Claude Code skill?

A skill is a markdown file with frontmatter (`name`, `description`, `user-invocable: true`) plus optional supporting files. When Claude Code starts a session it indexes every skill in `~/.claude/skills/`. When your request matches a skill's trigger phrases — or when you type the skill's slash command (e.g. `/clarify`) — Claude invokes it. From the agent's side, "invoking a skill" means loading the SKILL.md content and following its instructions.

Skills compose. One skill can chain to another by reading its file or calling its slash command. Several of the skills here chain to each other (e.g. `respond-html` calls `chrome-validate` for QA; `fresh-eyes` uses `delegate` to spawn verification subagents); the rest are standalone.

The eleven skills

Output & QA

Skill	Problem it solves	Standalone?
`respond-html`	When a chat reply has implicit structure (plan, audit, comparison, proposal), the chat format buries it. `respond-html` writes a self-contained HTML artifact with reading-optimized layout, inline ✅/💬/❌ reactions per section/callout/decision-block, and a one-click "copy feedback as markdown" button. The user reacts in the browser, pastes the markdown back to chat, the agent iterates.	⚠️ Chains to `chrome-validate` (bundled) for QA, `dataviz` (bundled) for data + process cont