Experimentalist — Design agent for Claude Code
Experimental design and execution specialist.
How to install Experimentalist
Installs to ~/.claude/agents/rhowardstone-claude-code-scientist-experimentalist.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/rhowardstone/Claude-Code-Scientist/HEAD/.claude/agents/experimentalist.md -o ~/.claude/agents/rhowardstone-claude-code-scientist-experimentalist.md Restart Claude Code, or start a new session, for it to be picked up.
What Experimentalist does
name: experimentalist description: Experimental design and execution specialist. Designs, implements, validates, and runs in-silico experiments with proper controls. Use when RQs require experimental evidence. model: opus
Role: Experimentalist
You design, implement, and run in-silico experiments to test hypotheses. You work in PHASES, with Research Director review between each.
Phases
Phase 1: DESIGN
- Review hypothesis to test
- Identify REAL data sources (APIs, databases,
Alternatives in Design
- Auto Claude — by AndyMik90 - Autonomous multi-agent coding framework for Claude Code (Claude Agent SDK) that integrates the 13.4k ★
- 1code — Best UI for Claude Code with local and remote agent execution 5.3k ★
- Omd A11y Auditor — Stage 0 deterministic gate of the eval pipeline 465 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Dev Frontend Developer
Frontend developer who builds user interfaces, implements responsive designs, and handles client-side function
Referral Architect
Designs and implements referral programs with reward mechanics, K-factor modeling, DB schema, API endpoints, U
Mechanical
Implement bounded, explicitly specified code changes that require no substantive design choice.
Frontend Gate
Very senior frontend engineer who gates a code diff BEFORE it lands on develop. Reviews the changed UI for the
Experimental Design Architect
Specializes in translating high-level research philosophy into concrete, falsifiable, and scalable experimenta
Statistician
Statistical rigor advisor: experimental design, signal-vs-noise, pressure-testing metric claims, sample-size a
Related Skills
Hypothesis
Generate research hypotheses and experimental designs based on data patterns
Hol Guard AWS Preflight
Fail-closed pre-execution HOL Guard skill for agents running aws CLI commands. Implements the aws/agent-toolki
Hazmat
Containment for AI agents - user isolation, sandboxed execution, network controls, backup/rollback. TLA+ verif