v0.6.0
Fuzzbucket v0.6.0 — Skills, Workflows, Audio Nodes, and a More Powerful Canvas
Fuzzbucket now connects more of the creative process in one workspace: teach it with Skills, start from reusable templates, bring research and brand media onto the canvas, and turn generations into editable output.
Fuzzbucket
What's New
Skills and Reusable Workflows
- Added a Skills library to discover, install, and use Skills in chat or create private Skills with your own references and examples.
- Added workflow templates with previews, compact inputs, Run/Stop controls, and expansion for deeper editing.
- Fuzzy can select the right Skill or template and shares clearer progress while working.
Creative Canvas and Media
- Added connected text, image, video, and first-class audio nodes with voice previews, model and voice selection, and canvas connections.
- Added drag-and-drop media, reusable brand assets across workspaces, and one-click placement of web research images without regenerating them.
- Added GPT-5.6 models and video controls that adapt inputs, durations, and credit estimates to the selected model and resolution.
- Added stronger image editing and version history, artifact attachments in chat, archive and restore, plus permanent and multi-select deletion.
Reliability, Pricing, and Sharing
- Long-running workflows now survive refreshes, reuse unchanged steps, avoid duplicate outputs, and show clearer progress, cancellation, and errors.
- Added yearly plans, direct plan changes, accurate generation pricing, safer credit handling, and clean original media for eligible paid usage.
Fixes & Improvements
- Fixed canvas state loss, destructive autosaves, duplicate nodes, stale generation states, and incorrect model inputs.
- Improved brand onboarding and deletion, dashboard and Content loading, uploads, mobile layouts, themes, localization, and Arabic RTL support.
