Skills agent-browser

📦

agent-browser

Name: agent-browser
Author: skillssh

Content revision r2 High Risk ⚙️ External commands🌐 Network access📁 Filesystem access

Automate Browser Tasks with Agent Browser

Manual web interaction slows research, testing, and data collection. This skill controls hosted browser sessions with structured navigation, actions, extraction, and visual capture.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "agent-browser" from https://skillstore.io/skills/skillssh-agent-browser.md and its manifest at https://skillstore.io/api/skills/skillssh-agent-browser/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/skillssh-agent-browser.md Signed manifest GET /api/skills/skillssh-agent-browser/manifest Signed lockfile GET /api/skills/skillssh-agent-browser/lockfile

Test it

Using "agent-browser". Inspect the contact page and identify the form controls.

Expected outcome:

Opened the contact page. Found name, email, message, and submit controls. Closed the session after inspection.

Using "agent-browser". Capture the approved product page and summarize its main sections.

Expected outcome:

Captured viewport and full-page images. The page contains overview, specifications, pricing, and support sections.

Using "agent-browser". Test the sign-up form with the supplied test account.

Expected outcome:

Submitted the test form once. The confirmation page loaded, and the browser session was closed.

Security Audit

High Risk

v5 • 9/7/2026 Open versioned report

Most static alerts are documentation syntax, placeholders, or standard shell patterns without command injection. Screen capture and upload are confirmed because hosted sessions can transmit sensitive content. Semantic review also found credential, cookie, rate-limit circumvention, and dependency integrity risks.

Files scanned

2,315

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (5)

High

SKILL.md:3

Screen capture upload

description: "Browser automation for AI agents via inference.sh. Navigate web pages, interact with e

The skill explicitly supports screenshots, video recording, and file upload through a hosted inference.sh browser. These features can transmit sensitive page or local-file content outside the user environment.

High

references/authentication.md:31-38 references/authentication.md:99-102 references/authentication.md:131-138

Sensitive Credentials Sent Through Hosted Browser Sessions

The authentication guidance sends account passwords and TOTP codes through inference.sh browser commands. Primary credentials can therefore cross an external service boundary.

The examples directly place password and generated TOTP values into hosted browser interaction requests. The surrounding text identifies these values as authentication secrets.

High

references/authentication.md:209-219

Session Cookies Printed to Command Output

The cookie extraction example reads document.cookie and prints the result. This can expose reusable session material in terminal logs or agent transcripts.

The example explicitly assigns document.cookie output to COOKIES and echoes it. This is direct disclosure of available session-cookie values.

High

references/proxy-support.md:85-127

Proxy Rotation Presented for Rate-Limit Avoidance

The proxy guide explicitly recommends rotating proxies for web scraping under a rate-limit avoidance heading. This can facilitate bypassing service controls.

The section title states rate-limit avoidance and the example rotates proxy endpoints across scrape targets. The circumvention intent is explicit.

Medium

SKILL.md:7 SKILL.md:17

Unpinned Third-Party Skill Installation

The setup directs users to install belt-sh/cli without an immutable version or commit. Future upstream changes could execute with the granted Bash tool access.

The installation command names a moving package reference and the prerequisite links to remote installation instructions. No version or commit pin is shown.

Risk Factors

⚙️ External commands (50)

🌐 Network access (50)

📁 Filesystem access (17)

references/authentication.md:5 references/authentication.md:263 references/commands.md:3 references/commands.md:59 references/proxy-support.md:5 references/proxy-support.md:272 references/session-management.md:5 references/session-management.md:136 references/snapshot-refs.md:5 references/video-recording.md:5 references/video-recording.md:239 templates/authenticated-session.sh:30 templates/capture-workflow.sh:28 templates/capture-workflow.sh:78 templates/capture-workflow.sh:81 templates/capture-workflow.sh:149 templates/form-automation.sh:23

Detected Patterns

Screen capture upload

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/skillssh-agent-browser/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/skillssh-agent-browser/security.svg)](https://skillstore.io/skills/skillssh-agent-browser?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/skillssh-agent-browser?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/skillssh-agent-browser/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/skillssh-agent-browser.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

skillssh. (2026). agent-browser security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/skillssh-agent-browser/audits/5

BibTeX citation

@techreport{skillssh-skillssh-agent-browser-2026,
  author = {skillssh},
  title = {agent-browser security audit report (audit version 5)},
  institution = {Skillstore},
  year = {2026},
  number = {5},
  url = {https://skillstore.io/skills/skillssh-agent-browser/audits/5},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "agent-browser security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
  - name: "skillssh"
date-released: "2026-09-07"
url: "https://skillstore.io/skills/skillssh-agent-browser/audits/5"
identifiers:
  - type: other
    value: "skillstore:skillssh-agent-browser:audit:5"
    description: "Skillstore immutable audit report identifier"

Compare variants

19 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Highest Skillstore Score

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

supercent-io Recommended

supercent-io-agent-browser

Skillstore Score 78

Evidence Confidence High

Skillstore usage 44

Updated

2026-09-09

101-skills

101-skills-agent-browser

Skillstore Score 74

Evidence Confidence Medium

Skillstore usage 4

Updated

2026-09-09

vercel-labs

vercel-labs-agent-browser

Skillstore Score 38

Evidence Confidence High

Skillstore usage 79

Updated

2026-09-09

qu-skills

qu-skills-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 42

Updated

2026-09-09

tul-sh

tul-sh-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 38

Updated

2026-09-09

am-will

am-will-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 37

Updated

2026-09-09

inference-sh-8

inference-sh-8-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 27

Updated

2026-09-09

skillssh Current

skillssh-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 18

Updated

2026-09-09

toolshell

toolshell-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 17

Updated

2026-09-09

inference-sh-9

inference-sh-9-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 17

Updated

2026-09-09

inferen-sh

inferen-sh-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 12

Updated

2026-09-09

inference-sh-skills

inference-sh-skills-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-09-09

inference-shell

inference-shell-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-09-09

inferencesh

inferencesh-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-09-09

inference-sh-0

inference-sh-0-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 11

Updated

2026-09-09

infsh-skills

infsh-skills-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-09-09

inf-sh

inf-sh-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-09-09

inference-skills

inference-skills-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 8

Updated

2026-09-09

halt-catch-fire

halt-catch-fire-agent-browser

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 4

Updated

2026-09-09

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Validate Web Journeys

Run repeatable navigation and form workflows, then capture page state for review.

Collect Structured Research

Open approved sources, extract visible text and links, and preserve screenshots.

Reproduce Support Issues

Record authorized browser steps and collect visual evidence for difficult user-interface problems.

Try These Prompts

Open and Inspect a Page

Open [URL], list the interactive elements, summarize the visible page, then close the browser session.

Complete a Test Form

Open [FORM URL], inspect current element references, fill the supplied test data, submit once, and report the result.

Capture Research Evidence

Visit these approved URLs: [URLS]. Extract relevant text and links, take full-page screenshots, cite each source, and close every session.

Debug a Stateful Journey

Reproduce [WORKFLOW] on [TEST SITE] with a disposable account. Re-snapshot after each change, capture failure evidence, and redact sensitive values.

Best Practices

Use only authorized sites, accounts, files, and proxy endpoints.
Re-snapshot after navigation or dynamic updates before using element references.
Close every session and disable recording when sensitive data may appear.

Avoid

Do not enter primary credentials, recovery codes, or production secrets into hosted sessions.
Do not use proxy rotation to bypass rate limits, access controls, or service policies.
Do not reuse stale element references after page state changes.