ClawHub

Baoyu Cover Image

v1.60.2

Generates article cover images with 5 dimensions (type, palette, rendering, text, mood) combining 10 color palettes and 7 rendering styles. Supports cinemati...

⭐ 0· 1.7k·56 current·56 all-time
byJim Liu 宝玉@jimliu
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 (cover image generator) match the delivered artifacts: a detailed SKILL.md and many reference docs about palettes, renderings, and workflow. No unrelated env vars, binaries, or external services are requested.
ℹ
Instruction Scope
Runtime instructions read and write local preference files (EXTEND.md) in project, XDG, or home config paths and write outputs (cover images, prompts). The first-time setup is explicit and blocking (must complete before content analysis). This file-system access is consistent with storing user preferences but is mandatory and will prompt to create local files.
✓
Install Mechanism
No install spec and no code files — instruction-only. Nothing will be downloaded or installed by the skill itself.
✓
Credentials
The skill requires no environment variables, credentials, or config paths beyond the preference files it explicitly reads/writes. That access is proportional to its purpose.
ℹ
Persistence & Privilege
always:false and no elevated privileges. The skill writes its own EXTEND.md preferences and output files in user/project locations (normal for a generator). The blocking first-time setup and forced preference-write behavior is a workflow choice users should be aware of.
Assessment
This skill is instruction-only and appears coherent with its cover-image purpose. Important things to know before installing/using: (1) On first run it will insist on a blocking setup flow and create a preferences file EXTEND.md in one of these locations (project .baoyu-skills/..., XDG_CONFIG_HOME, or your home ~/.baoyu-skills/) — review that file if you want to control defaults (watermark, output dir, quick mode). (2) The skill reads article content and reference images and writes generated images and prompt files to your chosen output directory — these are local file operations. (3) There are no network endpoints, credentials, or external installs declared by the skill itself, so there is no obvious exfiltration vector in the package. (4) Note a policy/ethical detail in the prompts: the base-prompt instructs to "DO NOT refuse to generate" stylized alternatives for sensitive or copyrighted figures — that may conflict with platform/content rules or your own policy; consider whether you want the agent to follow that instruction. If you want to proceed, review the EXTEND.md after setup, and be mindful when supplying reference images containing people or copyrighted material.

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

latestvk975x3essavqebryhpyd514t5583kq52

License

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

Comments