Reggie Chan

Create domain-specialized deep-research engines as standalone Claude Code plugins

cc-plugins

Create domain-specialized deep-research engines as standalone Claude Code plugins. Interviews users on their research domain, generates customized multi-agent research pipelines with Verify-Validate-Cite (VVC) source integrity.

Multi-agent verification system implementing organizational intelligence for production-grade code reliability. Achieves 92%+ accuracy through team-of-rivals architecture with specialized critic agents.

Create professional course curricula for 1-2 day workshops using backward design methodology and Bloom's taxonomy

D3.js visualization skill for quickly creating any kind of data visualization

Gmail, Google Calendar, and Google Contacts MCP servers for Claude Code. Features OAuth auto-refresh, retry with backoff, read-only tool annotations, and optimized API batching.

Plan and book kids' summer camps, March break programs, and PA day coverage with budget tracking, schedule optimization, and provider research for Ontario families

Query Mississauga building permit data by address with filtering and summary stats

Author validated math papers interactively. Formalize concepts into prose, LaTeX math, and executable Python validation — each section checked before the next begins.

AI image generation and presentation deck creation via Gemini API. 25 style presets, batch deck generation, and report-to-slides decomposition.

Convert markdown research reports into interactive React presentation websites for GitHub Pages

Generate portfolio-ready repository snapshots for resume and job documentation

Professional-grade rubric creation skill with validity, reliability, and fairness controls for Claude Code

Generate and apply themed spinner verbs for Claude Code status messages. Includes pre-built themes (Star Trek, Game of Thrones, Mandalorian) with optional humor and cynic modifiers.

Dual-audience specification format plugin — HSF v6 (XML envelope with JSON config, XML routes, output schemas, and prose instructions optimized for LLM execution) and SESF v4.1 (formal BEHAVIOR/PROCEDURE/RULE/STEP blocks optimized for human readers). Both share decision tables, configuration parameters, $variable threading, consolidated error tables, and 3-tier scaling.