Pinned Loading

1. wait-what-bench wait-what-bench Public

   Benchmark: Does an LLM notice a non-sequitur?

   Python

2. discovery-skill discovery-skill Public

   Guided discovery session skill for Claude Code — structured facilitation with first-principles decomposition

3. e2e-skill e2e-skill Public

   Claude Code skill for E2E testing — discover coverage gaps, author human-readable scenarios, and generate/run Playwright tests with AI.

4. sac-website sac-website Public

   JavaScript 5 2

5. autoeval autoeval Public

   Transform vague optimization problems into fully scaffolded autonomous experiment loops. Claude Code skill.

   Python

6. terminal-agent: plain English to Pow... terminal-agent: plain English to PowerShell command, Ctrl+G swaps the line in place

   1

   # terminal-agent

   2

   3

   Plain English in, shell command out. Nothing runs until you press Enter.

   4

   5

   ## Use