Each keeps its full history under its own directory. cowork-complete-guide's
static site is served at /cowork-complete-guide/ on the unified Pages site.
Also fixes stale miti99dev.github.io host references left from the org move.
Defines triggers, input requirements, sub-agent usage patterns, and
output formats for synthesizing insights across multiple documents.
Covers standard research synthesis, vendor comparisons, multi-perspective
analyses, and web + local source combinations. Includes example prompts.
Analyzes second location expansion, technology stack investment, and
espresso machine upgrade from Financial, Customer Experience, and
Operational Risk perspectives. Each decision synthesized into a concrete
recommendation with conditions and phased implementation plans.
Side-by-side comparison across pricing, lead time, service network,
reliability, resale value, and brand prestige. Includes individual
verdicts and key insight that capability differences between machines
are often overstated — decision hinges on service access, lead time,
and resale value.
Comprehensive analysis of the 2025-2026 specialty coffee market covering
6+ themes: independent shop advantages, technology as hygiene factor,
labor cost reset, sustainability imperatives, remote work geography shift,
and competitive strategy. Includes contradictions matrix, actionable
recommendations, and open strategic questions.
Runs file-processor skill on 5 inbox files: moves and renames to
processed/receipts, processed/invoices, processed/documents, and
processed/images using ISO date naming convention. Generates
outputs/processing-log-2026-03-09.md. Inbox is now empty.
Adds 5 realistic test files: Amazon receipt, freelance contract draft,
web design invoice, product sync meeting notes, and an app bug screenshot.
These serve as demo input for the file-processor skill.
Introduces a repeatable file management workflow: inbox for new files,
processed for organized output, outputs for Claude-generated reports,
and reference for read-only context. Includes a file-processor skill
with naming conventions and categorization rules. Updates README with
usage instructions.
Organizes 19 files into 7 topic folders using DONE/In Progress status labels
(Analytics, Budget, Marketing, Meetings, Menu & Recipes, Operations, Training).
Files renamed with ISO dates and descriptive names. Includes organization-summary.md.
Organizes 10 files from root into 4 subfolders (finance, operations,
marketing, planning) with descriptive names. Includes CHANGES.md
documenting the reorganization.
- Add .claude/commands/ideas.md: scans vault and CLI graph data to
produce tools to build, essays to write, connections to make,
and experiments to run
- Update my-commands.md with /ideas description and usage
- Add .claude/commands/challenge.md: pressure-tests beliefs by finding
contradictions and counter-evidence within vault notes
- Add .claude/commands/emerge.md: surfaces hidden patterns and unstated
ideas implied by the vault's collective content
- Update my-commands.md with descriptions and usage for both commands
- Add .claude/commands/connect.md: bridges two disparate vault topics
by tracing backlinks and surfacing unexpected connections
- Update my-commands.md with /connect description and usage
Adds the full initial course content including:
- START-HERE.md with navigation and teaching instructions
- 12 lesson scripts covering delegation, frameworks, research, automation, and more
- Scenario files for customer-feedback, done-framework, file-chaos, first-task,
large-corpus, receipts, and strategy-notes exercises
Add initial lesson structure for Vin Obsidian Workflows mini-lesson:
- CLAUDE.md: full lesson script and teaching instructions
- my-vault/: 10 interconnected Obsidian notes about Claude Code best practices
- .claude/commands/start-lesson.md: slash command to launch the lesson
- my-commands.md: documentation for custom commands
Define 9 features with requirements and test conditions covering auth,
reading experience, history tracking, search/discovery, ratings, comments,
notifications, and guest access. Includes UI and technical decisions
using Next.js + PostgreSQL.
Convert all 41 TypeScript source/config files in cc4e-course/quiz-project
(Next.js), gstack (Astro), and superpowers (Vite/React/Phaser) to plain
JavaScript with JSDoc type annotations, replacing tsconfig.json with
jsconfig.json (strict + checkJs) in each subproject. No behavior changes:
lint, typecheck, test, and build gates match their TypeScript baselines
exactly in all three subprojects. Update stack references in the affected
READMEs accordingly.
Replace bun.lock with package-lock.json and add package.json#allowScripts for
esbuild and sharp, the two native packages whose install scripts npm 11 gates.
Update README, RUNBOOK, TODOS and CLAUDE.md to npm commands. CLAUDE.md also
claimed ./setup requires bun; it does not any more.
Verified: astro check 0 errors, 39 tests pass, 4 pages build, clean npm ci.
The quiz app is a pure client-side Next.js page, so it static-exports with
output: 'export' and a basePath under the site; the lesson thumbnail is a
self-contained HTML file served as that directory's index.
Each web demo builds with its base path under /ai-coding-workflow-labs/,
a single deploy workflow assembles them with a root index, and the
per-directory workflows from the standalone repos are removed.