ResearchFlow V0.1.1 is the first archived open-source release of the project.

ResearchFlow is a local research knowledge base plus AI research assistant. It is designed to turn papers, analysis notes, and ideas into reusable research memory shared by Claude Code, Codex CLI, and other AI agents.

This release serves as the V0.1.1 archive baseline before the next-stage evolution of ResearchFlow toward a stronger evidence layer for automated research and a future ResearchWiki workflow.

What is included

• local research knowledge base structure centered on paperAnalysis/, paperPDFs/, paperCollection/, and paperIDEAs/
• shared skill library under .claude/skills
• bilingual documentation (README.md and README_CN.md)
• helper scripts for collection, download, analysis, maintenance, and Codex compatibility setup
• example knowledge base content for a motion-generation research direction

V0.1.1 changes

• added paperIDEAs/README.md to document the downstream output layer
• unified Chinese documentation naming from *_ZN to *_CN
• removed .codex/ from tracked repository state and kept it as local generated compatibility state
• cleaned up repository docs to better match the current ResearchFlow structure
• refreshed scripts/README.md to remove legacy ResearchWY wording
• fix: added missing STATE_CONVENTION.md with complete state definitions (analysis_mismatch, too_large)
• fix: updated README.md and README_CN.md to reference full state flow and STATE_CONVENTION.md

Notes

• .claude/skills is the single maintained source of truth
• .codex/ should be generated locally with python3 scripts/setup_shared_skills.py
• Obsidian is optional and not required for the core workflow
• this release is intended as an archive version rather than a polished final product

Attached archive

• ResearchFlow.V0.1.1.zip
• ResearchFlow.V0.1.1.zip.sha256

SHA256:
3f4039be606df987eebea1369873312cb553ba30d98fc7793f9939bbf93e8081

Release commit:

• 69d3130 — fix: complete STATE_CONVENTION.md with analysis_mismatch and too_large states