人生密码 / Rensheng Mima

Local-first Chinese metaphysics computation toolkit and web MVP.

人生密码是一个本地优先的中文命理结构化计算工具。它把八字、紫微斗数、文墨天机兼容文字盘解析、格局识别和融合报告拆成可测试的工程模块，并提供一个 React/Vite 与静态降级并存的本地网页入口。

This project is for cultural research, structured rule-engine experiments, and personal reflection. It is not medical, legal, financial, or psychological advice.

Live static demo: https://rucmeteor1025.github.io/rensheng-mima/

The GitHub Pages demo is a privacy-safe static frontend. It uses synthetic fallback summaries because GitHub Pages cannot run the Python rule engine. For real local computation, run python3 web/server.py.

项目亮点

• 可测试的规则引擎：八字、紫微斗数、格局检测、融合层和文墨兼容解析分别沉淀为可回归测试模块。
• 本地优先隐私模型：公开仓库不包含真实出生记录、姓名、生成报告、日志、token 或 agent state。
• 可运行 Web MVP：本地 API + 首页表单 + 免费摘要 + 价格页 + 隐私页 + 报告删除流程。
• 开源协作友好：CI、issue template、release checklist、贡献指南、隐私扫描脚本和路线图都已放在仓库内。

快速上手

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python3 scripts/xiashensuan.py 1990 1 15 14 --minute 30 --gender 男 --place 北京 --mode life

Run the local web MVP:

python3 web/server.py

Open http://127.0.0.1:8765.

Run the frontend dev server:

cd web/frontend
npm ci
npm run dev

产品预览

All screenshots below use synthetic/default demo input only.

Home	Analysis

能力范围

• BaZi engine with true-solar-time correction, ten-god analysis, weighted five-elements scoring, day-master strength, useful-god strategy, and luck-cycle summaries.
• Zi Wei Dou Shu engine using a C5FYC-style rule path, palace maps, major stars, support stars, transformations, and boundary-sensitive double-chart handling.
• Wenmo-compatible text parser for private local regression workflows.
• Pattern detectors for Zi Wei chart structures.
• Fusion layer for BaZi, Zi Wei, MBTI, zodiac, and blood-type modules.
• Local web MVP with free summary, mock checkout flow, privacy page, pricing page, and report deletion.

隐私边界

No real user birth records, names, generated reports, private notes, logs, tokens, or agent state are included in this public package.

The public xiashensuan_core/fixtures/wenmo_cases/manifest.json is intentionally empty. Maintainers can keep private fixture records locally under private_records/, which is ignored by Git.

Before publishing or merging a contribution, run the privacy scan and tests:

python3 -B scripts/wenmo_text_scan.py --record-dir private_records

前端开发

cd web/frontend
npm ci
npm run build

After building, restart python3 web/server.py; the server will serve web/frontend/dist/ first and fall back to web/static/ if no build exists.

测试

python3 -B scripts/ziweiwenmotests.py
python3 -B scripts/xiashensuantests.py
python3 -B scripts/ziwei_patterns_tests.py
python3 -B scripts/wenmo_text_scan.py --record-dir private_records

Private Wenmo fixture tests are skipped when no private fixtures exist.

贡献路线

Good first contribution areas:

• add synthetic public fixtures that do not come from real users
• improve parser edge-case coverage
• split output schemas into typed JSON contracts
• polish the local web MVP and frontend smoke tests
• expand privacy review checks for pull requests
• improve examples, screenshots, and docs

Open work is tracked in GitHub Issues. Please read CONTRIBUTING.md, PRIVACY.md, and SECURITY.md before submitting changes.

Repository Scope

Included:

• rule engines
• local web MVP source
• deterministic tests
• privacy and contribution docs
• synthetic screenshots and quickstart demo assets

Excluded:

• real case records
• generated reports
• private notes
• private agent files
• .openclaw or local runtime state
• node_modules
• frontend build artifacts

Roadmap

• Add synthetic public fixtures that do not come from real users.
• Split the engine into a proper Python package.
• Add typed JSON schema for chart outputs.
• Add GitHub Pages demo screenshots or static preview.
• Add CI coverage for backend smoke tests and frontend preview checks.
• Add privacy review checklist for contribution review.

License

MIT.