AI chats disappear into account histories. Research trails lose their context. Copy and paste strips away where information came from, and screenshots are difficult to search.
BrowserBeast™ turns browser pages, AI chats, research trails, documentation, and highlighted text into durable local Markdown or plain-text records—complete with source context, capture metadata, Evidence Envelope structure, and advisory privacy signals.
Built for people who need records... with receipts.
BrowserBeast preserves more than text. It records where a capture came from, when it was created, how it was captured, and which tool created it.
The result is a readable local artifact that can be stored, searched, reviewed, and used with AI tools without depending on a proprietary cloud archive.
- Full Render capture for readable page content
- Highlighted Text capture for precise selections
- Markdown and plain-text export
- Local Full Review for larger captures
- Evidence Envelope v0.2
- Source URL, timestamp, platform, browser, and capture-mode metadata
- Advisory privacy-signal detection
- Artifact-ID-based filenames
- Local-first, user-triggered operation
- No account, cloud service, or automatic background capture
BrowserBeast does not upload captured content to AI Underground. Captures remain under the user’s control.
Privacy detection is advisory, not automatic redaction. Always review an artifact before sharing it.
- Download
BrowserBeast-v1.0.zip. - Unzip it.
- Open
chrome://extensions. - Enable Developer mode.
- Click Load unpacked.
- Select the extracted
BrowserBeast-v1.0folder.
BrowserBeast™ is a Chrome Extension, Freeware, and Open Source under the MIT License.
Project home: https://ai-underground.ai/ Source: https://github.com/DuskStormcrow/BrowserBeast
Capture. Export. Enjoy.
BrowserBeast™ was developed through an ongoing collaboration between one human developer and multiple OpenAI models.
This was not a "generate an app from a prompt" project. Development consisted of months of iterative engineering conversations involving architecture, implementation, debugging, documentation, testing, release engineering, and product design.
GPT-5.5 served as the primary engineering and product collaborator throughout BrowserBeast's development. Responsibilities included:
- Product architecture and feature design
- Technical problem solving and implementation guidance
- JavaScript, TypeScript, HTML, and CSS development support
- Documentation and technical writing
- UI and UX refinement
- Testing strategy and bug analysis
- Release planning
- GitHub, Chrome Web Store, and Build Week preparation
GPT-5.6 working through Codex focused on repository quality and release engineering, including:
- Code review and implementation verification
- Repository auditing and cleanup
- Release validation
- Documentation review
- Packaging and public release preparation
- GitHub workflow recommendations
- Final quality assurance before publication
All product direction, implementation decisions, testing, final approvals, and releases remained under human control.
OpenAI's models acted as engineering collaborators, reviewers, teachers, and technical partners—not autonomous software authors.
BrowserBeast is the result of that collaborative workflow.