# claude-skills

Six skills that write and check Claude skills, slash commands and subagents: generate from requirements, audit structure, improve the result, design a CLI for helper scripts.

- plugin · `type=plugin` · Claude Code · ★ 276 repo stars
- Author: Samarth Gupta
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/gupsammy/Claudest/tree/HEAD/plugins/claude-skills
- Updated: 2026-10-29
- Security check: automated, 2026-11-16 (text and 6 script files) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/plugins/samarth-gupta-claude-skills

## What it does

The plugin covers the whole loop of building skills and subagents for Claude. `create-skill` asks about the objective, the trigger phrases and the inputs, then drafts `SKILL.md` and tests its description against prompts that should and should not fire it. `create-agent` does the same job for subagents.

Auditing splits in two. `repair-skill` and `repair-agent` check seven dimensions: frontmatter, execution modifiers, rules taught by example instead of stated outright, steps that belong in a script, verbosity, workflow clarity, and whether the directory holds the files it needs. `improve-skill` asks another question: does the skill give users the outcome they came for. Keep the pair apart, one fixes form, the other fixes results.

\`create-cli` designs the command-line interface for scripts that deterministic steps get extracted into. Once a skill has been created or improved, a bundled `skill-lint` agent runs the structural check on its own. There are no dependencies, and the skills run on whichever Claude model Claude is using.

## Install

### Claude Code (native platform)

**Plugin marketplace**
In the agent chat (not a shell):
```text
/plugin marketplace add gupsammy/Claudest
```
In the agent chat (not a shell):
```text
/plugin install claude-skills@Claudest
```

## When to use it
- generate a skill from written requirements instead of a blank file
- audit an existing skill and get a list of structural violations
- spot workflow steps that belong in a separate script
- design the CLI for a script a skill calls
- raise an existing skill to the outcome users actually need

## Example prompts
- Create a skill that pulls test results from `reports/` and writes a summary to `summary.md`
- Audit my deploy skill and list every structural violation
- Improve my release skill, it keeps skipping the changelog step

## Similar
- [agnix](https://affpapa.org/en/nevibecode/skills/agent-sh-agnix.md) — skill · ★ 493 repo stars
- [cmdxray](https://affpapa.org/en/nevibecode/mcp/cmdxray.md) — MCP server · ★ 5 repo stars
- [prumo](https://affpapa.org/en/nevibecode/mcp/prumo.md) — MCP server · ★ 5 repo stars
- [learn-taste](https://affpapa.org/en/nevibecode/skills/learn-taste.md) — skill · 13700 installs
- [project-min-evaluation](https://affpapa.org/en/nevibecode/skills/project-min-evaluation.md) — skill · 5700 installs
- [om-create-skill](https://affpapa.org/en/nevibecode/skills/om-create-skill.md) — skill · 4000 installs

## Other versions with the same name
- [claude-skills · surajhk](https://affpapa.org/en/nevibecode/plugins/claude-skills.md) — ★ 8 repo stars

From the same set: [gupsammy/claudest](https://affpapa.org/en/nevibecode/by/gupsammy/claudest.md) — 47 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
