LabscoMCP early access ↗
REPO PACKAGECOMMUNITY

degausai wonda

degausai · publisher132 GitHub starsgithub.com/degausai/wonda

LABSCO SUMMARY

Summary

wonda is one skill wrapping a single CLI that generates images, video, music, and speech, edits short-form video, and can post the result to five social platforms, billed against a paid wonda.sh account.

Generation covers images, video, text, music, speech, and multi-speaker dialogue. Editing is aimed squarely at TikTok/Reels-style output — animated captions, text overlays, splitting or merging clips, background removal, upscaling, lip-sync, and more, roughly two dozen operations in total. Publishing reaches Instagram and TikTok, including carousels. Layered on top of all of that is LinkedIn and X automation — search, profile and message reading, posting, liking, following — plus Reddit posting and commenting, scraping Instagram/TikTok/Reddit profiles and the Meta Ads Library for competitor research, and performance analytics for connected accounts.

All of it lives in one skill file; someone who only wants image generation installs the same thing as someone who wants LinkedIn outreach automation, because wonda does not split the two apart. An account at wonda.sh is required (wonda auth login), and every generation or edit draws down purchased credits, checkable ahead of time with wonda pricing estimate.

READ MORE

The social automation stores a session, not an API key.

wonda linkedin auth set and wonda x auth set capture a logged-in session for wonda to drive directly, rather than going through either platform's official developer API — a mechanism the README documents plainly but does not flag as running against those platforms' own automation rules, so that risk is left for whoever installs it to weigh.

No license file.

We found none in this repository, which is unusual for a tool meant to run inside an agent's shell holding live account credentials for five different platforms.