Search
Posts
Global Skill Schema and a Prediction Markets Vertical
Defining the Global Standard Skill Schema, a HOW-TO guide for perfect skills, ARD/PRD templates, a Polymarket Analyst skill, CI workflow, and a demo forecasting project. 106 files, 11,951 lines.
Readv1.2.0: Every Skill to 100% Compliance
24 commits of systematic compliance remediation: every skill from ~38-40% to 100%, sales language stripped from the README, non-compliant doc names archived.
ReadCEO Quickstart Guide and DeFi Sentinel Exploration
A 203-line quickstart guide written for one reader, plus a 6-document deep exploration of a DeFi prediction product that may never get built.
ReadImplementing Brand Consistency with CSS Variables: A Sponsor Page Redesign
How to migrate a sponsor page from inconsistent hardcoded colors to a cohesive brand design system using CSS variables, with real examples from Claude Code Plugins marketplace.
ReadSkills Infrastructure from Scratch: 8 SKILL.md Files, an Installer, and a Compliance Framework
Building the entire skills infrastructure for the Nixtla workspace from zero: 8 skill definitions, an installer package, demo project, compliance audits, and strategy documentation.
ReadThree Releases in One Day: Nixtla Goes Enterprise
Doc-Filing v3.0 compliance, Enterprise Plugin README Standard v1.0.0, and three version bumps — transforming the Nixtla repo from developer prototype to CEO-ready business showcase.
ReadNixtla BigQuery Forecaster: Cloud Functions on 200 Million Rows
Plugin #2 for the Nixtla workspace: a Cloud Function running AutoETS and AutoTheta on BigQuery public data — 200M+ Chicago taxi trip rows with keyless WIF deployment.
ReadGitHub Profile Refresh and Nixtla Review Readiness
Professional GitHub README rewrite, Nixtla Baseline Lab v0.7.0 review kit with mobile checklist, and a Perception dashboard README restructure.
ReadNixtla Baseline Lab: Phases 4 Through 8 in a Single Day
Five phases of the Nixtla Baseline Lab shipped in one day — setup automation, CI with golden task harness, PNG visualization, CSV support, and a TimeGPT showdown mode.
ReadNixtla Baseline Lab: Real statsforecast Baselines via MCP Server
The Nixtla Baseline Lab gets a working MCP server running SeasonalNaive, AutoETS, and AutoTheta baselines on M4 data with sMAPE and MASE evaluation metrics.
Read