Skip to content

Vision and Roadmap

Mustapha Zoghlami edited this page Nov 4, 2026 · 3 revisions

Vision and roadmap

Where it is going

  1. From flows to a library of skills. Every task you do becomes a replayable flow, then a shareable skill that needs no model. (Flows, Skills)
  2. Local models that learn from you. Qwen/Laya get better from your feedback and your own demonstrations, so routine pages cost your cloud agent fewer tokens. (Local-Models)
  3. A plugin ecosystem with visible permissions. Narrow tools, honest manifests, approval before install. (Plugins)
  4. More browsers and platforms. Firefox and Chrome today; Windows installer; tested macOS and Windows next; then other browsers, and mobile explored through the same protocol.

Each direction says what exists today and what comes next in docs/VISION.md, together with how BrowserForge differs from headless automation, extension bridges and LLM-driven reverse-API tools.

Principles that will not change

Local-first · the human approves · page content is data · no anti-bot evasion · honest output · open source (MIT).

Roadmap

The planned work is on the public BrowserForge roadmap board (Todo / In progress / Done) and tracked as GitHub issues labelled roadmap: open roadmap items. Highlights:

  • Code-sign the Windows installer (#4)
  • Test on real Windows and macOS machines (#5)
  • Dashboard v2: switch plugins on/off with approval (#7)
  • A shared skill library (#8)
  • First fine-tuned local decider (#9)
  • Safari and Android explorations (#13, #14)

Have an idea? Open a feature request or plugin idea.

Clone this wiki locally