domains / research / autonomous-research-optimization / autoresearch

autoresearch

Skill card

combined

The two skills share the name 'autoresearch' but define fundamentally different procedures. The first is a marketing/content optimization loop that generates variants of copy and scores them with a simulated 5-expert panel. The second is a scientific research orchestration skill that manages a two-loop architecture (inner experiment loop, outer synthesis loop) for autonomous AI research projects, including literature review and paper writing. They do not reproduce, shorten, or wrap each other; they are distinct workflows that happen to share a name.

Repositories

Other copies

Combined skill

# Autoresearch

This skill covers two distinct autonomous optimization workflows. Determine which mode is requested based on the user's intent.

## Mode 1: Content Optimization (Marketing)
Use when optimizing landing pages, email sequences, ad copy, or other conversion-focused content.

1. **Intake:** Parse the source content and identify optimizable elements (headlines, CTAs, etc.).
2. **Panel Setup:** Define a 5-expert scoring panel (e.g., CMO, Skeptical Founder, CRO, Copywriter, CEO).
3. **Optimization Loop:**
   - Generate 10 variants per element.
   - Batch-score all variants with the panel in a single API call.
   - Keep the top 3 variants.
   - Evolve the top 3 into new variants for the next round.
   - Repeat until a score threshold (e.g., 80/100) is met or 3 rounds are complete.
4. **Cross-Breeding:** Combine the best elements from each category into a final optimized version.
5. **Output:** Save the optimized content, experiment log, and optimization report.

## Mode 2: Autonomous Research Orchestration (Science)
Use when starting a research project, running autonomous experiments, or managing a multi-hypothesis research effort.

1. **Bootstrap:** Clarify the research question, search literature, and form initial hypotheses.
2. **Workspace Setup:** Create a structured directory for state, logs, findings, literature, code, data, and experiments.
3. **Two-Loop Execution:**
   - **Inner Loop:** Run rapid experiment iterations with clear optimization targets. Record results and learnings.
   - **Outer Loop:** Periodically synthesize results, identify patterns, update findings, and steer research direction.
4. **Finalize:** Write the research paper and produce a final presentation.

## Conflicts
- **ericosiu/ai-marketing-skills** focuses on simulated expert scoring of marketing copy variants.
- **brycewang-stanford/Auto-Empirical-Research-Skills** focuses on autonomous scientific research orchestration with a two-loop experiment/synthesis architecture.