skatkov/opencode-lightpanda — explained in plain English
Analysis updated 2026-05-18
Fetch JavaScript-heavy web pages inside OpenCode and get clean markdown back.
Extract structured JSON data from a web page for an AI coding workflow.
Replace the slower built-in WebFetch tool with a faster headless browser.
| skatkov/opencode-lightpanda | 0xkinno/neuralvault | 0xmayurrr/ai-contractauditor | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | hard | easy |
| Complexity | 2/5 | 4/5 | 2/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires installing the separate Lightpanda browser binary before the plugin will work.
opencode-lightpanda is a plugin for OpenCode, an AI coding assistant, that adds a faster and more capable way to fetch web pages. The author describes it as similar to the built-in WebFetch tool, but much stronger, calling it that tool on steroids. The plugin is built on top of Lightpanda, a lightweight headless browser. According to the README, Lightpanda runs about ten times faster than headless Chrome and uses about sixteen times less memory. Because it is a real browser rather than a simple page fetcher, it can load pages that rely heavily on JavaScript to build their content, something a basic fetch request often cannot handle. Once a page loads, the plugin returns clean markdown text with ads and clutter stripped out, so the assistant reading it gets readable content instead of raw, messy HTML. It can also return the page as structured JSON data or as a semantic tree, depending on what the calling tool needs. To use it, you first install Lightpanda itself, then install this plugin globally through OpenCode's plugin command. If OpenCode cannot find the Lightpanda executable automatically, you can point to it directly using an environment variable. Each time the tool is called it starts a fresh Lightpanda process, so no cookies or browser history carry over between calls, keeping every fetch independent and predictable. The project is written in TypeScript and uses Bun for local development, testing, and running checks. Developers who want to work on the plugin itself can also load their local checkout directly into OpenCode's configuration file instead of installing the published version. OpenCode also lets users control this tool and the standard WebFetch tool separately through permission settings, so one can be allowed while the other is denied. This is a narrow, practical developer tool aimed at people already using OpenCode who want faster, cleaner web page access for their AI coding workflows.
A faster, cleaner web-fetching plugin for OpenCode built on the lightweight Lightpanda browser, returning markdown, JSON, or a semantic tree instead of raw HTML.
Mainly TypeScript. The stack also includes TypeScript, Bun, OpenCode.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.