🧪
OpenClawFree

Agent Browser for OpenClaw: Ref-Based Browser Automation Candidate

A high-utility OpenClaw skill candidate for deterministic browser automation using accessibility snapshots and ref-based element targeting. Not yet tested by Runner AI Review.

May 28, 2026
Original
Awesome OpenClaw Skills category listOpen original externally↗
#openclaw#skill#agent#free#browser#automation#testing
Status & Access
Current access and latest update details.
Access
Free
Updated
Jun 1, 2026, 10:39 AM

LinkLoot AI review

Start skill only in isolation

Score: 63/100
Skill checked in isolation

Reviewed loot: Agent Browser for OpenClaw: Ref-Based Browser Automation Candidate

My take: this skill has solid evidence: the linked skill sources were fetched, scanned statically, checked for Python imports and dependency surface, and stored as sandbox artifacts.

User decisionVerify first

this skill has solid evidence: the linked skill sources were fetched, scanned statically, checked for Python imports and dependency surface, and stored as sandbox artifacts.

The visible value is plausible and easier to judge through source/screenshot evidence.Judges how careful a user should be when trying it: permissions, network use, dependencies, and hard warnings.
Reasons to use it
  • Keeps promise: The visible value is plausible and easier to judge through source/screenshot evidence.
  • Easy to try: Judges whether a normal user can repeat the first setup with reasonable effort.
  • Sources, external URL, and visible link/site signals were reviewed.
  • Install surface, scripts, dependencies, and suspicious patterns were checked statically.
Reasons to be careful
  • The runner found 5 place(s) that can start programs, use install scripts, or run code dynamically. For this loot: try it in a test environment first, do not use real tokens/cookies, then...
  • 5 spots mention credentials, browser sessions, root/admin mode, proxies, or similar access-sensitive behavior. This fits the tool category, but it means testing should happen with throwaw...
  • Public GitHub issues show user questions or improvement requests. Examples: #1397 Interrupted open can leave detached daemon / Chrome-for-testing state and stale sockets…; #1371 Orphaned...
  • Details currently unavailable.
Keeps promise84/100
Safe to try43/100
Easy to try76/100
Trust signals68/100
Worth following69/100
LLQI auditVerdict: RISKY
security & trust58/100
functionality & value78/100
quality & structure66/100
sources checked
code signals checked
snapshot optional
Dependency note

Automated AI review. Decision aid, not a safety guarantee. · 2026-06-01 04:50:34 UTC

What it does

Agent Browser is an OpenClaw community skill candidate for controlling web pages through a dedicated browser automation CLI. Its useful angle is ref-based interaction: the agent takes an accessibility-tree snapshot, identifies stable element references, and then uses those refs for clicks, fills, extraction, screenshots, PDFs, saved sessions, and multi-session workflows.

Pricing classification: free. Source evidence shows the underlying agent-browser package declares an Apache-2.0 license and the public repository exposes an Apache License file.

Who should use it

Use this candidate for review if you often need reliable browser workflows where CSS selectors are too brittle: multi-step forms, dynamic single-page apps, login-state reuse, parallel admin/user sessions, and structured extraction from web UIs. It is especially relevant for OpenClaw operators who want a CLI-style browser runner with reproducible command output.

Setup surface

The skill surface references a global agent-browser CLI and Chromium installation. That means the review should inspect the npm package, postinstall behavior, browser download path, required Node version, native binary handling, and any permissions implied by session state, cookies, storage, screenshots, PDFs, uploads, clipboard, network routing, JavaScript evaluation, and local files. No installation or execution has been performed on this Raspberry Pi.

Runner test plan

  1. Static scan: inspect the skill markdown, metadata, referenced commands, GitHub repository, package manifest, scripts, native binaries, and release artifacts without executing them.
  2. Dependency/install review: review npm package metadata, postinstall script behavior, Chromium download behavior, binary provenance, Node/Rust requirements, lockfiles if available, and license evidence.
  3. Prompt-injection/tool-poisoning review: check whether the SKILL.md or examples try to override agent instructions, request secrets, broaden file access, bypass approvals, or encourage unsafe execution.
  4. Sandbox execution: only after static approval, install and run in a disposable sandbox/VM with no real secrets, isolated home directory, restricted network as appropriate, and explicit logging of filesystem/network effects.
  5. Screenshot/video evidence: capture a simple navigation, accessibility snapshot, click/fill flow, screenshot/PDF output, and any command output so reviewers can verify behavior.
  6. Residual risks: document browser credential leakage, cookie/session persistence, arbitrary JavaScript evaluation, clipboard/file upload exposure, network interception misuse, native binary trust, and update-channel drift.

Risk notes

This Loot is a candidate, not a safety endorsement. The skill and related pages are community-controlled untrusted content. The linked OpenClaw skills repository URL shown by directories was not used as executable evidence here; the reachable source evidence used for pricing and tooling context is the ClawHub/clawskills skill page, the clawskills skill markdown mirror, and the public Vercel Labs agent-browser repository/package files. Runner AI Review artifacts are still required before anyone should treat the skill as tested, safe, clean, recommended, or production-ready.

Source links

Sources & links
References, demos, and supporting links.
Awesome OpenClaw Skills category listraw.githubusercontent.comPrimaryClawHub pageclawhub.aiClawskills listingclawskills.shSkill markdown source mirrorclawskills.shUnderlying agent-browser source repositorygithub.comPackage metadataraw.githubusercontent.comLicense evidenceraw.githubusercontent.com
Discussion

Sign in to join the discussion and vote on comments.

No comments yet. Start the discussion.
Keep exploring

More from this topic

More in OpenClaw