Issue Station
Sign In
Tools hereToolbox

claude-code - 💡(How to fix) Fix [FEATURE] Build Chrome Replacement [1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
anthropics/claude-code#48537•Fetched 2026-04-16 06:57:24
View on GitHub
Comments
0
Participants
1
Timeline
4
Reactions
0
Author
epicivyepicivy
Participants
epicivyepicivy
Timeline (top)
labeled ×3renamed ×1

Error Message

Browser automation is one of Claude Code's most powerful capabilities, but it's also the most fragile. Every user who hits a "browser extension not connected" error loses trust in the platform. Owning the browser layer turns a liability into a competitive advantage.

Fix Action

Fix / Workaround

Manifest V3 restrictions — Google is actively limiting what extensions can do, and further lockdowns are expected Extension policy changes — Google can delist or restrict the Claude in Chrome extension at any time Anti-automation design — Chrome is built for human users, not AI agents. Every automation capability is a workaround, not a supported use case Connection fragility — Native messaging, tab groups, and extension lifecycle are all failure points that generate constant support burden Proposal Anthropic already ships a desktop app. Bundle a lightweight browser engine (Chromium-based webview or similar) that:

Connects natively to Claude Code without an extension layer Provides first-class automation APIs instead of screen-scraping workarounds Can't be broken by Google policy changes Handles auth, cookies, and sessions in a way designed for AI-assisted workflows This doesn't need to replace Chrome for general browsing — just for the automation use cases Claude Code needs.

RAW_BUFFERClick to expand / collapse

Preflight Checklist

  • I have searched existing requests and this feature hasn't been requested yet
  • This is a single feature request (not multiple features)

Problem Statement

Anthropic-built browser to eliminate Chrome dependency for web automation

Proposed Solution

Summary Claude Code's browser automation is entirely dependent on Chrome via an MCP extension. This creates a fragile, unreliable foundation that Google can break at any time. Anthropic should build or bundle its own browser/webview purpose-built for AI-assisted web automation.

Problem A quick search of this repo tells the story — dozens of open issues about the Chrome extension not connecting (#45230, #40758, #37541, #35675, #41879, #37027). This isn't a bug to fix; it's an architectural dependency on a platform Anthropic doesn't control.

Specific risks:

Manifest V3 restrictions — Google is actively limiting what extensions can do, and further lockdowns are expected Extension policy changes — Google can delist or restrict the Claude in Chrome extension at any time Anti-automation design — Chrome is built for human users, not AI agents. Every automation capability is a workaround, not a supported use case Connection fragility — Native messaging, tab groups, and extension lifecycle are all failure points that generate constant support burden Proposal Anthropic already ships a desktop app. Bundle a lightweight browser engine (Chromium-based webview or similar) that:

Connects natively to Claude Code without an extension layer Provides first-class automation APIs instead of screen-scraping workarounds Can't be broken by Google policy changes Handles auth, cookies, and sessions in a way designed for AI-assisted workflows This doesn't need to replace Chrome for general browsing — just for the automation use cases Claude Code needs.

Why this matters Browser automation is one of Claude Code's most powerful capabilities, but it's also the most fragile. Every user who hits a "browser extension not connected" error loses trust in the platform. Owning the browser layer turns a liability into a competitive advantage.

Alternative Solutions

No response

Priority

Medium - Would be very helpful

Feature Category

MCP server integration

Use Case Example

No response

Additional Context

No response

extent analysis

TL;DR

Anthropic should consider bundling a lightweight browser engine, such as a Chromium-based webview, to replace the Chrome extension for automation use cases.

Guidance

  • Identify the specific requirements for the bundled browser engine, including native connectivity to Claude Code and first-class automation APIs.
  • Research existing lightweight browser engines, such as Chromium-based webviews, to determine the best fit for Anthropic's needs.
  • Evaluate the potential benefits of owning the browser layer, including reduced fragility and increased control over the automation experience.
  • Consider the potential impact on users who rely on Chrome for general browsing and ensure a smooth transition for automation use cases.

Example

No code snippet is provided as this issue is focused on high-level architecture and design decisions.

Notes

The proposed solution requires significant development and testing efforts, and may involve trade-offs between functionality, performance, and maintainability. It is essential to carefully evaluate the costs and benefits of this approach before proceeding.

Recommendation

Apply workaround: Bundle a lightweight browser engine to replace the Chrome extension for automation use cases, as this approach addresses the identified risks and limitations of the current architecture.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#ssr#chain error#conversation history#tool integration

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING