GLM-V-PDF-to-WEB
v1.0.0Convert a PDF (research paper, technical report, or project document) into a beautiful single-page academic/project website with a structured outline JSON. T...
⭐ 0· 172·0 current·0 all-time
by@zai-org
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (PDF→website) matches the code and instructions: python-based scripts convert PDF pages to images, crop images, and save HTML/outline files. The only minor mismatch is that SKILL.md lists the use of curl as a system tool but the registry metadata's required binaries list includes only 'python' (curl is expected but not declared).
Instruction Scope
Runtime instructions are narrowly scoped to downloading a PDF (if a URL), converting pages to images, reading pages to produce a structured outline.json, delegating cropping to a subagent, and saving the final HTML. The SKILL.md explicitly requires using a fresh subagent (Agent tool) for cropping and using the provided crop.py script. This is consistent with the task but means page images and crops will be handed to a separate agent context.
Install Mechanism
No automated install/spec is provided (instruction-only), and bundled scripts are small and local. Dependencies are standard Python packages (PyMuPDF, Pillow) installed via pip as documented — no downloads from unknown third-party URLs or archive extraction steps are present.
Credentials
The skill requests no credentials or special env vars. It will read local PDF files (or download a user-supplied URL to /tmp), create directories under the workspace and /tmp, and write outline.json, crops, and index.html. Because cropping is delegated to fresh subagents, image data (the PDF content) will be exposed to those agent contexts — a privacy consideration if the PDF is sensitive.
Persistence & Privilege
The skill does not request permanent/always-on presence, does not modify other skills or global agent configs, and uses only normal filesystem writes in the workspace and /tmp. Autonomous invocation is allowed (default) but is not combined with any other broad privileges.
Assessment
This skill appears to do what it says: it converts PDFs into images, plans sections, crops images, and writes an index.html and outline.json under your workspace. Before installing/run it: 1) Be aware it will write files under {WORKSPACE}/web/... and may download a URL you provide into /tmp. 2) It requires pip packages (pymupdf, pillow) — installing them will access PyPI. 3) Cropping is delegated to fresh subagents, so the PDF page images will be processed by additional agent contexts (avoid using with confidential PDFs). 4) The SKILL.md expects curl to be available (not declared in required binaries); ensure curl is present or pass local PDFs. 5) If you want extra assurance, inspect the three scripts (pdf_to_images.py, crop.py, generate_web.py) locally — they are short and readable — before running.Like a lobster shell, security has layers — review code before you run it.
latestvk973q6nnbpzwfa2f2gfrnzhm6d83vys0
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🌐 Clawdis
Binspython