Hey π
LLM-based review with Chain-of-Thought and Few-Shot prompting, VLM mode that analyzes figure layout and equation aesthetics from page snapshots, plus prompt injection detection for PDF integrity β this goes well beyond a simple "summarize my paper" tool. The side-by-side comparison interface for testing multiple prompts on the same manuscript is a nice touch for prompt engineering.
I built a tool on socialistic.ai that turns any skill into a shareable link. You send the link, the other person opens it and uses your skill. No configuring API keys, no setting up Deepseek connections, no deploying Cloudflare workers β they just open the link and upload their PDF.

Try it here
Send this link to a labmate before they submit and they get structured strengths/weaknesses/suggestions in minutes β no setup, no API bills on their end.
If you're interested, I'd love to explore a collaboration β like revenue sharing based on usage, so your work generates income.
Happy to chat β shesonglin@tinkerland.ai
Feel free to close this if not relevant π
Hey π
LLM-based review with Chain-of-Thought and Few-Shot prompting, VLM mode that analyzes figure layout and equation aesthetics from page snapshots, plus prompt injection detection for PDF integrity β this goes well beyond a simple "summarize my paper" tool. The side-by-side comparison interface for testing multiple prompts on the same manuscript is a nice touch for prompt engineering.
I built a tool on socialistic.ai that turns any skill into a shareable link. You send the link, the other person opens it and uses your skill. No configuring API keys, no setting up Deepseek connections, no deploying Cloudflare workers β they just open the link and upload their PDF.
Send this link to a labmate before they submit and they get structured strengths/weaknesses/suggestions in minutes β no setup, no API bills on their end.
If you're interested, I'd love to explore a collaboration β like revenue sharing based on usage, so your work generates income.
Happy to chat β shesonglin@tinkerland.ai
Feel free to close this if not relevant π