What problem does it solve?
This Skill eliminates the manual, error-prone work of turning an empirical paper into structured, reusable research components by extracting variables, datasets, models, mechanisms, identification, robustness, and heterogeneity into a consistent wiki.
Core Features & Use Cases
- End-to-end paper ingestion: Converts a local PDF/TEX (or a pre-prepared source) into wiki pages for the full empirical workflow, not just a summary.
- Evidence-grounded extraction: Prioritizes extracting operational details (e.g., variable construction, identification strategy, robustness batteries) and avoids inventing unsupported facts by marking unreported items.
- Graph-enabled knowledge linking: Generates
wiki/graph/edges.jsonl to connect papers to variables, datasets, models, and key research relations with confidence and evidence.
- Use case: When you read a new accounting/finance/econ paper that operationalizes the same construct differently, you can ingest it so the new variant lands alongside prior variants in the corresponding variable pages.
Quick Start
Ask the skill to ingest a local PDF into the empirical wiki with a topic hint, for example: empirical-ingest "<local-pdf-or-tex-path>" --topic "耐心资本与 ESG".