Pinned Loading

1. AI-Firmware AI-Firmware Public template

   Self-syncing Claude Code project template: rule kernel, 34 skills, committed memory library, session hooks. Improvements flow both ways between the template and the projects it spawns.

   PowerShell 1

2. savetokens savetokens Public

   Token-saving techniques for Claude Code and Codex, published as a component datasheet. Measured before-and-after figures, refreshed daily. Source for savetokens.tips

   HTML 1

3. Truenote Truenote Public

   RAG for business | Citations or nothing | Support teams need an answer they can verify

   TypeScript 1

4. willaicite willaicite Public

   Free, deterministic GEO audit that scores whether AI answer engines (ChatGPT, Perplexity, Claude, Google AI Overviews) can retrieve, read, and quote your website

   TypeScript 1

5. good-codex-subagents good-codex-subagents Public

   Smart Codex Subagents

6. tracebench tracebench Public

   Replayable evals for AI coding agents, built from my own developer sessions: does the agent verify claims, own mistakes, and self-correct under pressure? Deterministic, reproducible scoring with ev…

   HTML