Releases: MazzaWill/neo4j-python-pandas-py2neo-v3
Release list
v0.4.0 - Neo4j Knowledge Graph Skill
This release adds an open-source Codex/agent skill for Neo4j knowledge graph work.
Added:
- skills/neo4j-knowledge-graph/ with SKILL.md and OpenAI agent metadata.
- References for spreadsheet-to-graph modeling and safe Cypher/vector-index/GraphRAG patterns.
- profile_table.py utility for profiling CSV/Excel files before graph modeling.
- Tests for the profile_table.py utility.
Changed:
- Linked the skill from English and Chinese README files.
- Extended maintenance checks to compile and test the skill utility.
- Added GitHub topics for graphrag, codex-skill, and ai-agents.
Validation:
- Skill validator passes.
- Modern example tests pass.
- Skill utility tests pass.
- Python compile checks pass.
v0.3.1 - Project Positioning Polish
This release improves the project positioning without changing the repository slug.
Changed:
- Updated the README display title to Excel to Neo4j Knowledge Graph.
- Kept the historical repository slug neo4j-python-pandas-py2neo-v3 for continuity and existing links.
- Reworked the English and Chinese README introductions to frame the project as both a legacy py2neo v3 example and a modern Neo4j GraphRAG/vector-search example.
- Updated GitHub repository description to match the clearer positioning.
No runtime behavior changed.
v0.3.0 - Modern Neo4j GraphRAG Example
This release adds a modern Neo4j example based on the original invoice Excel project.
Added:
- examples/modern_invoice_graphrag/ with English and Simplified Chinese docs.
- Official neo4j Python driver integration through a CLI.
- Neo4j vector index and db.index.vector.queryNodes search Cypher.
- Deterministic local embeddings for keyless dry-runs and CI.
- Optional neo4j-graphrag[openai] requirements for production GraphRAG embeddings.
- CSV sample data that runs without pandas or external services.
- Offline unit tests for invoice normalization, embeddings, Cypher, and CLI helpers.
The legacy py2neo v3 example remains unchanged. This release adds a current Neo4j 5+/2026 GraphRAG/vector-search path without breaking existing users.
v0.2.2 - README Language Split
This release cleans up the project documentation for a more professional open-source entry point.
Changed:
- Reworked README.md as an English-first project entry point.
- Moved the full Simplified Chinese documentation to README.zh-CN.md.
- Added language switch links between the English and Simplified Chinese docs.
- Removed mixed-language duplication from the main README while keeping Chinese documentation available for existing users.
This is a documentation-only release.
v0.2.1 - OSS Maintenance Metadata
This release fills in the standard open-source maintenance metadata needed for a healthier public repository.
Added:
- MIT license.
- CONTRIBUTING.md, SECURITY.md, SUPPORT.md, and CODE_OF_CONDUCT.md.
- GitHub issue templates for bug reports, compatibility questions, and sample-data requests.
- Pull request template and CODEOWNERS.
- Lightweight GitHub Actions syntax check.
- README links for governance, support, security, and contribution paths.
This keeps the project positioned as an actively maintained legacy py2neo v3 / Neo4j 3.x educational example while dependency modernization remains tracked in #23.
v0.2.0 - Maintenance Restart
This release marks the restart of active maintenance for neo4j-python-pandas-py2neo-v3.
Highlights:
- Documented the current maintenance status and roadmap.
- Added compatibility notes for the original Python 3.6.5 / Windows 10 / Neo4j 3.x / py2neo 3 environment.
- Added quick-start guidance for Invoice_data_Demo.xls, local path setup, and Neo4j credentials.
- Added CHANGELOG.md for future release tracking.
- Started historical issue triage, closing resolved sample-data and compatibility questions while leaving the remaining reproducible compatibility issue open for follow-up.
This remains a legacy py2neo v3 educational/example project. Modern Python, pandas, Neo4j, and py2neo compatibility work is tracked in the README roadmap.