Releases·Tool Release
Dev releases open-source browser-only tool that gives OpenAI-compatible agents a stable chat UI in one click
An engineer shared an MIT-licensed, browser-only tool that takes any OpenAI-compatible agent URL and generates a stable, polished chat UI in one click for quick testing. It displays spans, turns, duration, and token stats side-by-side. The author built it to avoid flaky existing options like Hermes' dashboard and LangChain's test UI.
13 engagement1 sourceSun, Sep 27, 2026, 09:18 AM
AI Overview
- MIT-licensed tool runs entirely in the browser, takes any OpenAI-compatible agent URL, and shows spans, turns, duration, and token counts side by side
- Developer cites flaky, noisy alternatives in Hermes' dashboard and LangChain's test UI, plus poor performance stats, as motivation for building it
Related
Releases·Tool Release
browser-use/browser-use 0.13.13 ships OpenAI browser-tool driver after 0.13.12 PyPI cancellation
Browser Use shipped an OpenAI browser-tool driver in versions 0.13.12 and 0.13.13, backed by Browser Use Actor. The integration exposes 18 default browser tools and six optional actions usable with local browsers, Browser Use Cloud, or borrowed browser sessions. 0.13.13 supersedes 0.13.12, whose PyPI publication was cancelled before approval.
Browser Use shipped an OpenAI browser-tool driver in versions 0.13.12 and 0.13.13, backed by Browser Use Actor. The integration exposes 18 default browser tools and six optional actions usable with local browsers, Browser Use Cloud, or borrowed browser sessions. 0.13.13 supersedes 0.13.12, whose PyPI publication was cancelled before approval.
Updated Thu, Oct 15, 2026, 11:50 PM — 0.13.13 supersedes 0.13.12 after the latter's PyPI publication was cancelled, adding two integration fixes.
0 engagement2 sourcesgithubThu, Oct 15, 2026, 11:37 PM
Releases·Tool Release
I wanted an in-app browser and custom API keys inside Google Antigravity, so I built BetterGravity (Open Source). Here's how it works:
A developer built BetterGravity, a free MIT-licensed client mod for Google Antigravity, to fix friction points like alt-tabbing for testing, quota caps, and no UI customization. It patches Antigravity's loopback-served UI at runtime and hooks Chrome DevTools Protocol to give the coding agent an in-app browser. The mod adds BYOK with Gemini keys, a redesigned Willow UI, and desktop pixel pets.
A developer built BetterGravity, a free MIT-licensed client mod for Google Antigravity, to fix friction points like alt-tabbing for testing, quota caps, and no UI customization. It patches Antigravity's loopback-served UI at runtime and hooks Chrome DevTools Protocol to give the coding agent an in-app browser. The mod adds BYOK with Gemini keys, a redesigned Willow UI, and desktop pixel pets.
2 engagement1 sourceredditWed, Sep 16, 2026, 12:55 AM
Products·launched
Veduta – personal AI agent with a generative UI, running as a server daemon
Veduta is a personal agent inspired by terminal-based agents like Hermes/OpenClaw, but wrapped in a graphical interface. An LLM agent runs as a daemon on a server, tracking information for the user and presenting visual feedback without repeated prompting. It targets individuals who want persistent background monitoring instead of asking the same questions again.
Veduta is a personal agent inspired by terminal-based agents like Hermes/OpenClaw, but wrapped in a graphical interface. An LLM agent runs as a daemon on a server, tracking information for the user and presenting visual feedback without repeated prompting. It targets individuals who want persistent background monitoring instead of asking the same questions again.
1 engagement1 sourcehackernewsWed, Oct 14, 2026, 01:29 PM
Products·building
Show HN: Building AI agents client-side JavaScript
An exploration of running AI agent loops entirely in the browser with vanilla JavaScript instead of server-side Python frameworks. The author is building a demo and a guide covering browser-specific constraints like CORS, API keys, visual state, and remote tool calls. It targets developers who want low-infrastructure-cost agents, including fully offline setups.
An exploration of running AI agent loops entirely in the browser with vanilla JavaScript instead of server-side Python frameworks. The author is building a demo and a guide covering browser-specific constraints like CORS, API keys, visual state, and remote tool calls. It targets developers who want low-infrastructure-cost agents, including fully offline setups.
1 engagement1 sourcehackernewsThu, Sep 3, 2026, 09:34 PM
Products·building
OpenWand: AI cowork interface with context fetching and prompt storage
OpenWand is an LLM interface designed to optimize AI cowork experiences by eliminating manual context copying and repetitive prompt writing. It fetches context in one click, stores custom prompts, and integrates AI into the app's look and feel, supporting any model.
OpenWand is an LLM interface designed to optimize AI cowork experiences by eliminating manual context copying and repetitive prompt writing. It fetches context in one click, stores custom prompts, and integrates AI into the app's look and feel, supporting any model.
4 engagement2 sourcesredditTue, Aug 18, 2026, 10:48 PM