Browser API

Headless browser rendering.

When To Use This API

The Browser API is for agent workflows that need public, platform-native data without writing a custom collector for every task. Use it when a prompt needs evidence from live social records, when a Skill needs repeatable enrichment, or when product code needs the same contract an agent discovered through MCP.

This section lists 4 live Browser operations from the canonical OpenAPI spec. Each operation page documents the route, method, parameters, response envelope, pagination behavior when available, and the error vocabulary shared across UnifAPI.

• Use Browser operations when an agent needs public records from this platform.
• Call the documented endpoints through MCP or HTTP and keep the response envelope consistent.
• Combine these records with other UnifAPI APIs for research, monitoring, and enrichment workflows.

Agent Contract Notes

Agents usually discover these endpoints with the hosted UnifAPI MCP server, inspect the operation schema, then call only the records needed for the current task. Developers can call the same endpoints over HTTP with an API key, which makes the path from exploratory agent run to production integration direct.

Every successful response uses a stable envelope with a request id, data payload, optional pagination cursor, and billing metadata. Errors use the same closed vocabulary across platforms, so retry logic, user-facing explanations, and cost reporting can stay consistent even when a workflow spans multiple public-data sources.

When you are choosing an operation, start from the public identifier the user already has: a profile username, post id, channel id, subreddit name, search term, or URL-derived slug. Prefer the narrowest endpoint that answers the question, then expand only when the agent needs more evidence. This keeps record usage predictable, makes citations easier to explain, and avoids turning a focused research task into an expensive crawl.

Operations

Html

• POST /browser/html — Render a page to HTML — Render a URL in a headless browser and return the fully rendered HTML after JavaScript execution — use it to read meta tags, structured data, and content that only appears after client-side rendering.

Links

• POST /browser/links — Extract links from a page — Render a URL in a headless browser and return every link on the page as absolute URLs, including links injected by JavaScript.

Markdown

• POST /browser/markdown — Render a page to Markdown — Render a URL in a headless browser (running JavaScript) and return the page as clean Markdown — ideal for content analysis, summarization, and feeding pages to language models.

Screenshot

• POST /browser/screenshot — Capture a page screenshot — Render a URL in a headless browser and capture a screenshot, returned as a base64-encoded image.