C4T4

S Tier Skills — Development skill for Claude Code

Development community

Only the Claude skills and agents that are actually worth installing.

How to install S Tier Skills

This entry records only its repository, not the path inside it, so there is no exact command to give. Open C4T4/s-tier-skills and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What S Tier Skills does

Only the Claude skills and agents that are actually worth installing. Curated hard, kept short.

Alternatives in Development

  • Crypto Research Command — Think hard and execute comprehensive cryptocurrency research by calling all crypto analysis agents in parallel 3.4k ★
  • Fast Jev Compaction — Claude Code plugin that replaces the compaction summary with Jev decisions: every tool call and result is scor 425 ★
  • Claude Code Mcpinstall — Easy guide to installing Claude Code MCPs globally on your machine 239 ★

README

S-Tier Skills

Only the Claude skills that are actually worth installing.

S superpowers · security-audit-skill · humanizer · i-have-adhd
A
B
C
D
F the other ~10,000

Yes, only one row has anything in it. That is the entire point.

A skill gets a spot by showing the same prompt run with and without it. If you can't see the difference in five seconds, it goes in the bottom row with everyone else.

Contents

Engineering

[superpowers](https://github.com/obra/superpowers)

Spec → plan → execute, with the verification step it can't skip.

**You:** *Add rate limiting to the API*

Without With
Writes a middleware. Says "Done! ✅". It is untested and the limit is wrong. Asks 6 questions → writes spec.md → 9-task plan → runs in a git worktree → each task test-first → halts at task 4 because the test genuinely fails.

The win: it can't tell you it's finished when it isn't.

Security

[security-audit-skill](https://github.com/cloudflare/security-audit-skill)

Every finding goes to a second agent whose only job is to disprove it.

**You:** *Audit this repo*

Without With
32 "potential vulnerabilities". You spend 3 hours finding out 28 are noise. findings.json: 4 confirmed, 9 need validation, 19 discarded — a fresh agent dis