From research-factory
Track regression experiments in _EXPERIMENTS.md. Use after every regression run to log specification, results, and sign consistency. Prevents losing track of which specification produced which result.
How this skill is triggered — by the user, by Claude, or both
Slash command
/research-factory:experiment-registry Regression results to logRegression results to logThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- After EVERY regression run
Look for docs/_EXPERIMENTS.md. If missing, create with header:
# Experiment Registry
| ID | Date | Spec | DV | Key IV | N | R-sq | Key coeff (SE) | Sign match? | Table file |
|----|------|------|----|--------|---|------|----------------|-------------|------------|
| 1 | 2024-03-15 | Main OLS | ret_excess | ai_score | 12,345 | 0.045 | 0.025** (0.010) | YES | output/tables/C5_main.tex |
Read the registry to understand baseline results and check for consistency.
npx claudepluginhub xuxiguo/research-factory-claude --plugin research-factoryGuides reception of code review feedback: verify before implementing, avoid performative agreement, push back with technical reasoning when needed.
Design skill for brand identity, design tokens, UI styling, logo generation (55 styles, Gemini AI), corporate identity programs (50 deliverables), presentations (Chart.js), banners (22 styles), SVG icons (15 styles), and social photos across multiple platforms.