Heron

About

Heron is a super-intelligent AI agent that lives in your browser. Just describe what you want — Heron figures out which tools and agents to use, executes the steps, and brings back the results. No tab switching. No copy-pasting between
  tools.

🤖 MULTI-AGENT ORCHESTRATION
Heron delegates tasks to specialized agents for GitHub, Jira, Azure, SharePoint, and more. Ask a question in plain language and the Orchestrator routes it to the right place automatically.

🌐 BROWSER AUTOMATION
Heron can read, click, type, and navigate any web page — including internal tools with no API. Extract tables, fill forms, click buttons, and capture screenshots on any tab.

🧠 LONG-TERM MEMORY
Heron remembers facts about you, your team, and your work across sessions. It automatically builds a knowledge graph of entities, relationships, and findings from every run.

🔄 WORKFLOWS
Chain multiple agents into reusable workflows. Built-in workflows include PR review (fetch diff → check CI → analyze → post review) and daily standup (GitHub PRs + Jira sprint → priority list).

⏰ SCHEDULER
Run any task automatically on a cron schedule, as a monitor (poll until a condition is met), or as a one-time delayed task — all without keeping the browser open.

✍️ FORM PREFILL
Heron reads any form on any page and proposes values based on your context, chat history, and past fill history. Works with plain HTML, React, and Vue forms.

🖱️ RIGHT-CLICK ACTIONS
Select any text, right-click, and run Summarize, Translate, Explain, Page to Markdown, or any custom action you define — results appear inline or in the side panel.

🔌 BRING YOUR OWN PLATFORM
Connect any internal tool via OpenAPI spec, MCP server, or by scanning a page with DeepScan. Every endpoint and tool becomes available to the Orchestrator automatically.

⚙️ FULLY CONFIGURABLE
Every capability is visible and controllable in Settings. Choose your preferred AI provider and model, set per-tool approval gates, override system prompts, and more.

  HOW TO GET STARTED
  1. Install Heron and click the toolbar icon
  2. Go to Settings → AI Model and enter your API key
  3. Open the Side Panel or Chat window
  4. Start typing — describe anything you want to do