ClawHub

prompt-zs

v1.0.1

专业AI绘画与视频创作提示词生成助手。当用户需要生成AI绘画提示词、Midjourney/Stable Diffusion提示词、AI绘画描述、图像生成提示词时使用。也适用于文生视频提示词、图生视频提示词、视频创作描述、Runway/Pika视频提示词等场景。支持生成JSON格式提示词、标签化提示词、摄影风格描述...

⭐ 0· 136·0 current·0 all-time
by@bozoyan
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description (prompt generation for AI images/videos) match the provided SKILL.md, examples, and included helper script. The only helper script implements the 'Expert 1' JSON generator, which is a reasonable, proportionate auxiliary; other modes are handled by the SKILL.md instructions and example files.
✓
Instruction Scope
SKILL.md gives explicit, narrow instructions about format, triggers, and the decision flow. It does not instruct reading arbitrary system files, accessing environment variables, or sending data to external endpoints. It does suggest calling the local script for basic prompt generation (local-only).
✓
Install Mechanism
No install spec is present (instruction-only skill with a small local script). There are no downloads, external package installs, or archive extraction steps that would write arbitrary code to disk.
✓
Credentials
The skill declares no required environment variables, no credentials, and no config paths. The included Python helper uses only standard libs (json, sys) and takes input from the command line—proportionate to its purpose.
✓
Persistence & Privilege
Flags are default (always: false, agent can invoke autonomously). The skill does not request permanent presence or attempt to modify other skills or system-wide settings.
Assessment
This skill appears internally consistent and limited to generating and formatting prompts. Before installing, you may want to: (1) confirm you are comfortable with the SKILL.md's strict automatic-trigger wording (it says the agent must use this skill when certain keywords appear); (2) inspect the small included Python script if you run code locally (it only formats JSON from a command-line argument); and (3) avoid providing any unrelated credentials or private files when using the skill. No network endpoints, downloads, or secret requests were found.

Like a lobster shell, security has layers — review code before you run it.

latestvk97cm2d50884x372bpkgk7dsk9833b0g

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments