# Agent Browser Protocol Deterministic browser automation protocol for Claude, Codex, OpenCode, and MCP-style agent workflows. ## Agent Decision Summary - Risk level: elevated - Source confidence: medium - Recommended workflows: Browser automation, Coding agent workflow, Connector or protocol layer - Permission surface: browser, shell/files, external services - Agent JSON: https://www.openagent.bot/plugins/agent-browser-protocol.agent.json ## Summary Agent Browser Protocol is an open-source browser automation project that exposes deterministic browser actions for agent tools. It is aimed at developers who want a cleaner control surface between coding agents and a real browser. ## Guide Agent Browser Protocol is an open-source protocol layer for deterministic browser automation. ### What it is It exposes browser actions to AI agents through a clearer interface than raw ad hoc control. ### Why it matters Browser automation is becoming a common agent capability, and teams need a controllable boundary between the agent and the browser. ### How it works Evaluate it on a local web app first, then inspect action logs and failure recovery before connecting production accounts. ### FAQ - Is Agent Browser Protocol open source? - Yes. The repository is listed under the BSD-3-Clause license. - Who should use it? - Developers building browser tools for AI coding agents or MCP-style workflows should evaluate it. ## What It Does It exposes browser actions to AI agents through a clearer interface than raw ad hoc control. ## How To Evaluate Evaluate it on a local web app first, then inspect action logs and failure recovery before connecting production accounts. ## Why It Matters As more agents operate browsers, the interface between the agent and the browser becomes infrastructure. A protocol layer can make browser actions easier to inspect, replay, and secure. ## Best For - Developers wiring browser control into coding agents - Teams evaluating MCP-style browser tools - Builders who want deterministic browser automation primitives ## Not For - Non-technical users looking for a hosted automation app - Teams that only need passive page screenshots ## What It Actually Does - Protocol-shaped browser control: The project focuses on a deterministic interface for browser automation. - Why it matters: Protocols make browser capabilities easier to expose to multiple agents without rewriting every integration. - Agent-tool compatibility: It is positioned for Claude, Codex, OpenCode, and MCP-style workflows. - Why it matters: A shared browser control layer can reduce glue code across agent environments. - Inspectable automation: Deterministic actions are easier to log, replay, and constrain than free-form browser control. - Why it matters: Inspection is important before agents touch authenticated or business-critical web apps. ## Typical Use Cases - MCP browser tool: Expose browser actions to an agent through a controlled tool surface. - Agent QA: Let coding agents open and inspect local apps during frontend development. - Replayable workflows: Capture browser steps in a way that is easier to debug after failure. ## How It Compares - When to choose Agent Browser Protocol: Compare it with nearby plugins by looking at hosting model, integration surface, license, and whether the official docs show the workflow you need. ## Fit Matrix - Browser automation: strong. Agent Browser Protocol has multiple signals for browser automation, including matching tags, capabilities, category, or positioning. Required check: Run one non-sensitive website task and inspect clicks, waits, retries, and changed URLs. - Coding agent workflow: strong. Agent Browser Protocol has multiple signals for coding agent workflow, including matching tags, capabilities, category, or positioning. Required check: Run a small repository change and inspect the diff, tests, and rollback path. - Connector or protocol layer: strong. Agent Browser Protocol has multiple signals for connector or protocol layer, including matching tags, capabilities, category, or positioning. Required check: Connect one low-risk service, then inspect schemas, auth scope, errors, and logs. - Evaluation and observability: partial. Agent Browser Protocol has at least one signal for evaluation and observability, but should be checked against a real task before adoption. Required check: Add one repeatable test case and confirm results can run again in review or CI. - Reusable skill workflow: partial. Agent Browser Protocol has at least one signal for reusable skill workflow, but should be checked against a real task before adoption. Required check: Run one skill end to end and check whether it produces evidence or structured output. - Local or private AI stack: weak. Agent Browser Protocol is not primarily positioned for local or private ai stack in the current metadata. Required check: Verify hardware requirements, data path, storage, and whether all calls stay in your environment. ## Evidence - verified: Agent Browser Protocol is listed as open source. Source: License metadata: BSD-3-Clause - verified: Agent Browser Protocol has a recorded GitHub repository: theredsix/agent-browser-protocol. Source: Resource facts and GitHub source link. - inferred: Agent Browser Protocol supports these recorded deployment modes: cloud. Source: OpenAgent decision signal metadata. - inferred: Agent Browser Protocol is tagged with mcp, browser automation, protocol capabilities. Source: OpenAgent capability taxonomy. ## Missing Checks - Dedicated docs link is missing. - Repository freshness has not been recorded. ## Next Actions - Inspect repository: https://github.com/theredsix/agent-browser-protocol ## Facts - Category: plugins - Resource type: plugin - Open source: yes - License: BSD-3-Clause - Last verified: 2026-06-09 - GitHub repo: theredsix/agent-browser-protocol - GitHub stars: 471 ## Capabilities - mcp - browser-automation - protocol ## Structured Use Case Tags - browser-agent - developer-workflow ## Getting Started - Open the GitHub repository: https://github.com/theredsix/agent-browser-protocol ## Links - GitHub: https://github.com/theredsix/agent-browser-protocol ## Structured Outputs - JSON: https://www.openagent.bot/plugins/agent-browser-protocol.json - Markdown: https://www.openagent.bot/plugins/agent-browser-protocol.md - Agent JSON: https://www.openagent.bot/plugins/agent-browser-protocol.agent.json - Canonical: https://www.openagent.bot/plugins/agent-browser-protocol