How brycewang-stanford/auto-empirical-research-skills works
AERS is a catalog of agent-facing empirical-research skills, presented as a human-interruptible workflow from research idea to submission. The README documents the end-to-end stages; source excerpts also show a numeric benchmark scorer, a rubric-based skill evaluation harness, and a specification-search tool for auditing p-hacking. Workflow links below reflect the README’s documented sequence; implementation-level links are limited to relationships visible in the supplied excerpts or source index.
Main entry points: search.py Specification search · 33-Galaxy-Dawn-claude-scholar Topic refinement · 68-research-productivity-skills Literature review · 67-econfin-workflow-toolkit Data acquisition
Components 19
Research workflow
- Topic refinement
skills/33-Galaxy-Dawn-claude-scholar - Literature review
skills/68-research-productivity-skills - Data acquisition
skills/67-econfin-workflow-toolkit - Identification strategy
skills/50-brycewang-aer-skills/templates/python - Estimation modeling
demo-notebooks/statsPAI_pipeline_5.2.py - Robustness audit
skills/73-brycewang-p-hacking-skills/scripts/phack - Tables and figures
skills/50-brycewang-aer-skills/templates/python - Writing and review
skills/35-bahayonghang-academic-writing-skills - AI-text reduction
skills/48-de-AIGC-skills/scripts/provenance_scrub.py - Journal submission
skills/67-econfin-workflow-toolkit/paper-submission
Skill capabilities
- Agent skill catalog
skills
Evaluation and rigor
- Rubric evaluation
eval-harness/run_evals.py - Automated checks
eval-harness/lib/checks.py - Benchmark scorer
aers_score/cli.py - Exam discovery and grading
aers_score/exam.py
Analysis tools
- Specification search
skills/73-brycewang-p-hacking-skills/scripts/phack/search.py - Analysis design grid
skills/73-brycewang-p-hacking-skills/scripts/phack/grid.py - Search procedures
skills/73-brycewang-p-hacking-skills/scripts/phack/procedures.py
Other components
- Researcher or agent
Connections 15
- Researcher or agent → Agent skill catalog: requests work
README.md - Agent skill catalog → Topic refinement: offers skills
README.md - Topic refinement → Literature review: refines into review
README.md - Literature review → Data acquisition: informs data search
README.md - Data acquisition → Identification strategy: supports design
README.md - Identification strategy → Estimation modeling: guides estimation
README.md - Estimation modeling → Robustness audit: feeds audit
README.md - Robustness audit → Tables and figures: informs reporting
README.md - Tables and figures → Writing and review: supports writing
README.md - Writing and review → AI-text reduction: continues through
README.md - AI-text reduction → Journal submission: prepares submission
README.md - Rubric evaluation → Automated checks: runs checks
eval-harness/run_evals.py - Benchmark scorer → Exam discovery and grading: loads exam
aers_score/cli.py - Specification search → Analysis design grid: uses grid
skills/73-brycewang-p-hacking-skills/scripts/phack/search.py - Specification search → Search procedures: runs procedure
skills/73-brycewang-p-hacking-skills/scripts/phack/procedures.py