📦 Zen Koan Daily — 每日禅宗公案
v0.1.0每日推送一则禅宗公案，附水墨风插图与TTS朗读音频，并生成详细讲义：出处、背景、解读与修行方式，助你在忙碌生活中随时参悟禅机。
0· 82·1 当前·1 累计
by 下载技能包
最后更新
2026/4/6
安全扫描
OpenClaw
可疑
medium confidenceThe skill mostly matches its stated purpose (daily koan + image + TTS) but contains several inconsistencies and sloppy/incomplete instructions that could cause unexpected behavior or require unrelated credentials/tools—review before installing or running.
评估建议
This skill generally does what it says (generate koans, an image prompt, and TTS), but there are several things to check before use:
- Clarify Gemini usage: README and generate_image.py mention Google Gemini and suggest exporting GEMINI_API_KEY, but the registry metadata does not declare this credential. If you do not want to provide a Gemini key, confirm how images are generated (the shipped script only prints a prompt/command).
- Cross-skill exec: generate_image.py prints a command that call...详细分析 ▾
⚠ 用途与能力
Name/description align with the code: Python scripts generate koans, image prompts, and TTS audio. However the README and script comments reference Google Gemini (imagen-3.0) and instruct setting GEMINI_API_KEY, yet the skill metadata declares no environment variables/credentials. generate_image.py prints a command that runs a script in a different skill path (~/clawd/skills/yumfu/scripts/generate_image.py) rather than invoking an image API itself—this cross-skill dependency is unexpected and not justified in metadata.
⚠ 指令范围
SKILL.md and scripts instruct the agent to run local commands, write files under $HOME/.openclaw and /tmp, and to execute uv run commands via the exec tool. generate_image.py prints an exec command which runs a different skill's script (yumfu) — this delegates image generation outside the skill's own code. SKILL.md also mentions optional web_search without declaring that tool; the LLM is authorized to dynamically generate koans (fine) but the instructions give broad discretion ('LLM generates from classical sources') which could cause it to hallucinate or fabricate translations/attributions. These behaviors stay within the functional domain but include cross-skill execution and vague web-search guidance.
✓ 安装机制
No install spec (instruction-only plus small scripts) — lowest install risk. Required binaries listed (python3, edge-tts) match the scripts. README suggests pipx install edge-tts and exporting GEMINI_API_KEY for images but there is no automated download/install step in the registry metadata.
⚠ 凭证需求
Registry declares no required env vars, but README and code comments reference Google Gemini and suggest exporting GEMINI_API_KEY for images. That credential is not declared in metadata. The skill attempts to use only local tools (edge-tts) and user filesystem, so additional cloud credentials would be disproportionate unless you accept the README's optional Gemini usage—this mismatch should be clarified.
✓ 持久化与权限
always: false and no special privileges requested. The skill writes output under the user's home (~/.openclaw/media/outbound/zen-koan) and updates a local references/progress.json — standard for a user-facing skill. It does not request system-wide changes or modify other skills' config, though it instructs running code from another skill path which may assume that other skill is present.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv0.1.02026/4/6
Initial release: Dynamic LLM-powered Zen koan generation with ink wash art and TTS audio
● 无害
安装命令
点击复制官方npx clawhub@latest install zen-koan-daily
镜像加速npx clawhub@latest install zen-koan-daily --registry https://cn.longxiaskill.com镜像同步中