Saad Shahd
Saad Shahd is an individual software engineer and independent developer, not an organization. His work focuses on making distributed systems reliable and recoverable through architectural separation, domain-specific languages, state machines, event sourcing, and thoughtful human/AI collaboration.
At a Glance
- Professional and elite sports organizations
- Sports broadcasters, analysts, media, betting, and gaming operators
- Fintech and financial-services product and editorial teams
- Software engineering teams building distributed systems
- +1 more
AI Tools by Saad Shahd
(1)moo.md
Claude Code Plugins for Agents
Discussions
No discussions yet
Be the first to start a discussion about Saad Shahd
Latest News
The Engineer's Anxiety at the Penalty Kick: Shahd wrote about accountability and comprehension when coding agents make more engineering decisions, and described the moo.md workflow he built to keep ownership with the human.
One Flew Over the Context Window: Shahd published a critique of AI coding-agent workflows and explained the origins of the hope Claude Code plugin and its focus on systematic thinking, human ownership, and one focused change at a time.
Products & Services
Open-source Claude Code plugins and portable agent skills, including hope, hold, hunch, and sound. It makes agents work at the user's pace by turning intent, architecture, verification, routing, and comprehension into explicit interactive steps. It can also install skills for Codex, Cursor, Copilot, Gemini CLI, opencode, Windsurf, and Goose via skills.sh or the repository's installation flow.
A system for concurrent live sports-data collection using configuration-driven domain rules, state machines, event graphs, streaming infrastructure, replayable events, automated validation, and parallel enrichment. Shahd describes it as scaling to more than 5,000 collectors across sports.
Sports-data collection tools that replaced manual video clicking with computer-vision-assisted positioning and context-aware keyboard workflows, while enforcing valid event sequences and supporting features such as goalkeeper reaction tracking.
Market Position
As an individual engineer, Shahd positions his work around reliability, recoverability, and architectural separation rather than simply eliminating failure. moo.md is differentiated from autonomous coding-agent products by deliberately preserving human ownership: it asks interactive questions, previews choices, keeps decision work separate from tactical execution, and treats comprehension as a product requirement. His systems work is differentiated by combining configuration-driven domain logic with state machines, event streams, and replayable history.
Target Markets
- Professional and elite sports organizations
- Sports broadcasters, analysts, media, betting, and gaming operators
- Fintech and financial-services product and editorial teams
- Software engineering teams building distributed systems
- Developers and technical teams using AI coding agents
- Real-time sports event collection and validation at broadcaster scale
- Sports performance analysis, scouting, player recruitment, and live data operations
- Fintech editorial and content systems where regulatory or market rules change faster than engineering deploys
- Developer tooling for intentional, accountable use of coding agents
- Distributed systems that need replayability, failure recovery, and independently evolving domain rules