← Back to Skills Marketplace
spyfree

Ziwei Doushu

by Li Xin · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ⚠ suspicious
745
Downloads
2
Stars
2
Active Installs
11
Versions
Install in OpenClaw
/install ziwei-doushu
Description
Professional Ziwei Doushu consultation skill with offline, Beijing-standard calculation rules. Use when the user wants a polished Ziwei report from birth dat...
Usage Guidance
This skill appears to be a legitimate offline Ziwei chart generator, but there are important inconsistencies you should resolve before installing: (1) SKILL.md says 'offline only' yet the packaged scripts import third-party libraries (iztro_py, iztro, possibly cairosvg). If those libraries are not already present, installing them will require network access. (2) There is no install specification or packaged dependencies — you should confirm whether the runtime environment already includes iztro_py/iztro and any image backends, or be prepared to install them from PyPI/NPM from trusted sources. (3) The Python and Node scripts use different command-line argument names (e.g., --time vs --time-index); test the 'dual' engine mode and the CLI flags to ensure consistent behavior. If you cannot confirm bundled dependencies or are uncomfortable allowing the agent to install packages from the network, treat the skill as unusable until dependencies are explicitly provided or an install spec from trusted release sources is added. There is no sign of network exfiltration or secret access in the code, so the primary issues are dependency/install ambiguity and CLI mismatches rather than malicious behavior.
Capability Analysis
Type: OpenClaw Skill Name: ziwei-doushu Version: 1.0.2 The ziwei-doushu skill bundle is a legitimate tool for generating traditional Chinese astrology reports. The code in scripts/ziwei_chart.py and scripts/ziwei_engine_js.mjs implements complex calculation logic using the 'iztro' library, featuring robust input validation (e.g., date/time parsing, gender normalization) and safe SVG generation with proper XML escaping. While the Python script contains a minor bug (a missing 'subprocess' import required for the optional JS engine fallback), there is no evidence of malicious intent, data exfiltration, or prompt injection. The bundle adheres to its stated 'offline only' scope and lacks any suspicious network or file system activity.
Capability Assessment
Purpose & Capability
The name/description (Ziwei Doushu astrology) aligns with the included Python and Node scripts which call iztro/iztro_py to compute charts and produce JSON/optional images. The code artifacts and reference docs are consistent with the stated product (chart facts + interpretation).
Instruction Scope
SKILL.md states 'Offline only; no network dependency' and recommends running the Python script directly, but the Python script attempts to import iztro_py and raises an error asking the user to 'pip install iztro-py' if missing. The Node script imports 'iztro'. There is no install spec explaining whether these libraries are bundled or must be fetched, so in practice installing dependencies would likely require network access, contradicting the offline claim. Additionally, the SKILL.md recommended CLI (--time as HH:MM) does not map directly to the Node script which requires --time-index and different argument names, creating ambiguity in dual/JS engine usage.
Install Mechanism
No install specification is provided even though the scripts import third-party libraries (iztro_py for Python, iztro for Node, and optionally cairosvg for chart rendering). That means the skill will either fail if those packages are not preinstalled, or require the user/agent to install packages from PyPI/NPM at runtime (network). The lack of an explicit, trusted install source or vendor package increases operational ambiguity and risk.
Credentials
The skill declares no required environment variables or config paths and the code does not read environment variables, system credentials, or arbitrary files. There is no evidence of requests for unrelated secrets or broad system access.
Persistence & Privilege
The skill does not request 'always: true' and does not appear to modify agent/system configuration. It runs as a transient script and does not persist credentials or enable permanent presence.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install ziwei-doushu
  3. After installation, invoke the skill by name or use /ziwei-doushu
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
Refine storefront copy: clearer consultation framing and stronger chart-facts vs interpretation messaging for ClawHub discovery.
v1.0.1
Improve appeal and packaging: cleaner professional positioning, clearer fact-vs-interpretation framing, and removal of local virtualenv artifacts from the published package.
v1.4.4
Patch 1.4.4: upgraded offline chart aesthetics (card-based palace layout, central summary panel, higher-resolution JPG), and changed default chart output to JPG while keeping failure-safe fallback.
v1.4.3
Patch 1.4.3: update public skill description to explicitly state fully-offline mode, no headless/browser dependency requirement, and optional chart output.
v1.4.2
Patch 1.4.2: enforce offline-only chart pipeline (auto/cairosvg), remove headless browser dependency path, strip external URLs from methodology metadata, and keep chart output optional/failure-safe.
v1.4.1
Patch 1.4.1: removed online chart backend options, switched to offline-only chart backends (auto/headless/cairosvg), made image output optional by default, and added graceful fallback so report generation continues even if chart rendering is unavailable.
v1.4.0
Release 1.4.0: added high-quality Ziwei chart image generation with website-style rendering backends (ziwei-local offline bundle, ziwei-pub online, headless/cairosvg fallback), switched default reports to include JPG chart automatically, improved chart clarity (higher DPR), and updated docs for backend behavior.
v1.3.0
Major refresh 1.3.0: upgraded to framework-first Ziwei output (facts + classical interpretation signals) without fixed advice; added py/js/dual engine validation with normalized Chinese mutagen labels; improved report structure and SVG chart output; documented minimum dependency versions (iztro-py>=0.3.4, iztro>=2.5.7).
v1.2.0
Professional upgrade: added ming palace Sanfang-Sizheng view, yearly Si-Hua mapping, and stronger yearly anchor output.
v1.1.0
Enhanced: yearly horoscope anchors (--year), decadal listing support, richer structured output and markdown report.
v1.0.0
Initial release: iztro-py based charting with longitude correction and structured markdown/json output.
Metadata
Slug ziwei-doushu
Version 1.0.2
License MIT-0
All-time Installs 2
Active Installs 2
Total Versions 11
Frequently Asked Questions

What is Ziwei Doushu?

Professional Ziwei Doushu consultation skill with offline, Beijing-standard calculation rules. Use when the user wants a polished Ziwei report from birth dat... It is an AI Agent Skill for Claude Code / OpenClaw, with 745 downloads so far.

How do I install Ziwei Doushu?

Run "/install ziwei-doushu" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Ziwei Doushu free?

Yes, Ziwei Doushu is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Ziwei Doushu support?

Ziwei Doushu is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Ziwei Doushu?

It is built and maintained by Li Xin (@spyfree); the current version is v1.0.2.

💬 Comments