Skip to content
View stephenywilson's full-sized avatar

Block or report stephenywilson

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
stephenywilson/README.md

Stephen Y. Wilson

Founder of Catalayer — building practical open-source AI tools for market intelligence, financial news reasoning, model evaluation, and safer AI-assisted development.

I focus on software that helps people turn noisy information into clearer research, safer automation, faster workflows, and better operational decisions.


Open Source Tracks

My open-source work is organized around two main tracks:

Track Focus
Market Intelligence AI-assisted market research, financial news reasoning, signal evaluation, and decision workflows
AI Agent Tooling Safer Claude Code, Codex, Cursor, Cline, MCP, and AI coding-agent workflows

Featured Systems

System Projects What it does
MarketLayer by Catalayer MarketLayer Open-source AI market research console for US stocks, built around news, signals, watchlists, and decision workflows
News2Signal Stack News2SignalLab, News2SignalBench, SignalEvalRunner Local-first evaluation stack for testing whether AI systems can turn financial news into structured market-relevant signals
AI Agent Tools MCP-Doctor, RepoBlackbox, AgentSkillKit Diagnostics, safety layers, repo audits, reusable task packs, and workflow tools for AI-assisted development

Focus Areas

Area What I Build
Market Intelligence News analysis, watchlists, market context, research signals, and AI-assisted market workflows
Financial News Reasoning Tools and benchmarks for turning financial news into structured events, sentiment, and market-relevant signals
AI Evaluation Benchmarks, datasets, runners, reports, and leaderboards for testing model behavior and signal quality
AI Agent Infrastructure MCP diagnostics, local-first AI workflows, provider configuration, reusable task packs, and workflow automation
AI Coding Safety Repo protection, audit reports, guardrails, task benchmarks, and safer workflows for AI coding agents
Sourcing Automation Supplier comparison, RFQ workflows, landed cost analysis, procurement templates, and vendor decision systems

About Catalayer

Catalayer is a practical intelligence layer for market analysis, sourcing automation, and decision workflows.

The open-source projects here are lightweight, transparent versions of tools and ideas connected to the broader Catalayer ecosystem.

Catalayer AI is the reasoning and analysis layer behind the platform, designed to support structured analysis, financial news understanding, market context, and workflow intelligence.


Upcoming Open-Source Model

Catalayer AI is preparing its first open-source model: Catalayer AI Mini 500M.

This is a compact news-analysis model developed through Catalayer's own training pipeline and designed for financial news understanding, event interpretation, sentiment reasoning, and market-context analysis.

The model is intended to be an independently controlled open-source release from Catalayer AI — not a wrapper around a third-party chatbot. Larger private models, proprietary datasets, and production intelligence systems will remain separate from the open-source release.


Open Source Direction

I am building a focused collection of small, useful, polished open-source tools around:

  • AI-assisted market research
  • financial news reasoning and evaluation
  • provider-agnostic model evaluation
  • MCP diagnostics and AI tool troubleshooting
  • safer AI-assisted development workflows
  • repo audits, workflow safety layers, and reusable agent skills
  • sourcing and procurement workflows
  • local-first AI applications
  • practical decision systems for independent builders and small teams

The goal is simple: tools that solve real problems, not demos that only look good.


Links

Pinned Loading

  1. MarketLayer MarketLayer Public

    MarketLayer by Catalayer — open-source AI market research console for US stocks, with news, signals, watchlists, and decision workflows.

    TypeScript 1

  2. News2SignalLab News2SignalLab Public

    Complete local lab for financial news-to-signal evaluation, reports, static leaderboards, and synthetic market-news streams.

    Python

  3. MCP-Doctor MCP-Doctor Public

    Diagnose, inspect, and safely configure MCP servers for Claude, Cursor, VS Code, and AI agents.

    TypeScript

  4. RepoBlackbox RepoBlackbox Public

    Lightweight safety, evaluation, and workflow layer for Claude Code, Codex, Cursor, and other AI coding agents

    TypeScript 1

  5. agent-task-watcher agent-task-watcher Public

    Local-first CLI for inspecting what AI coding agents changed before you trust the result.

    TypeScript

  6. GitHub-AI-Code-Inspector GitHub-AI-Code-Inspector Public

    Inspect AI-generated GitHub repos and PR diffs for risk signals directly inside GitHub.

    TypeScript