์Šคํ‚ฌ browser-fingerprint-audit
๐Ÿ“ฆ

browser-fingerprint-audit

์ฝ˜ํ…์ธ  ๋ฆฌ๋น„์ „ r1 ์ค‘๊ฐ„ ์œ„ํ—˜ โš™๏ธ ์™ธ๋ถ€ ๋ช…๋ น์–ด๐ŸŒ ๋„คํŠธ์›Œํฌ ์ ‘๊ทผ๐Ÿ“ ํŒŒ์ผ ์‹œ์Šคํ…œ ์ ‘๊ทผ

Audit Browser Fingerprint Consistency

Contradictory browser and network signals can reveal automation or spoofing. This skill runs liarjs checks and explains the resulting consistency score.

์ง€์›: Claude Codex Code(CC)
Maintain this Skill? Claim it ๐Ÿ“Š 65 ์ ์ ˆํ•จ

๋‚ด ์—์ด์ „ํŠธ๋กœ ์„ค์น˜

์ด ์š”์ฒญ์„ ์—์ด์ „ํŠธ์— ๋ณต์‚ฌํ•˜์„ธ์š”. ๊ณต์‹ Skill ํŽ˜์ด์ง€์™€ ๋งค๋‹ˆํŽ˜์ŠคํŠธ๊ฐ€ ํฌํ•จ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.

์—์ด์ „ํŠธ ์š”์ฒญ
Review the Skillstore skill "browser-fingerprint-audit" from https://skillstore.io/skills/liarjsdev-browser-fingerprint-audit.md and its manifest at https://skillstore.io/api/skills/liarjsdev-browser-fingerprint-audit/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

์—์ด์ „ํŠธ๋Š” ๊ณ„์†ํ•ด์„œ ๊ณ„ํš์„ ํ‘œ์‹œํ•˜๊ณ  ๋ณด์•ˆ ์ •์ฑ…์—์„œ ์š”๊ตฌํ•˜๋Š” ํ™•์ธ์„ ์š”์ฒญํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.

์—์ด์ „ํŠธ๊ฐ€ ์ฝ๊ธฐ ์‰ฌ์šด ๋ฆฌ์†Œ์Šค

AI ์—์ด์ „ํŠธ, ํฌ๋กค๋Ÿฌ, ์Šคํฌ๋ฆฝํŠธ๊ฐ€ ์ „์ฒด ํŽ˜์ด์ง€ ๋Œ€์‹  ์ •๋ฆฌ๋œ ์ปจํ…์ŠคํŠธ๋งŒ ํ•„์š”ํ•  ๋•Œ ์ด ๋งํฌ๋ฅผ ์‚ฌ์šฉํ•˜์„ธ์š”.

ํ…Œ์ŠคํŠธํ•ด ๋ณด๊ธฐ

"browser-fingerprint-audit" ์‚ฌ์šฉ ์ค‘์ž…๋‹ˆ๋‹ค. Audit a temporary browser and summarize only failed checks.

์˜ˆ์ƒ ๊ฒฐ๊ณผ:

  • Score: 82 of 100. Classification: Suspicious.
  • Failed check: navigator.webdriver. The automation flag is enabled.
  • Failed check: timezone mismatch. The browser timezone differs from the IP-derived timezone.

"browser-fingerprint-audit" ์‚ฌ์šฉ ์ค‘์ž…๋‹ˆ๋‹ค. Run an offline audit and explain the coverage.

์˜ˆ์ƒ ๊ฒฐ๊ณผ:

The audit completed using local browser signals only. It evaluated 32 checks and skipped eight cross-layer network checks.

"browser-fingerprint-audit" ์‚ฌ์šฉ ์ค‘์ž…๋‹ˆ๋‹ค. Explain why a headless browser received a low score.

์˜ˆ์ƒ ๊ฒฐ๊ณผ:

The low score reflects observable headless and automation signals. It measures profile consistency and does not predict a specific website response.

๋ณด์•ˆ ๊ฐ์‚ฌ

์ค‘๊ฐ„ ์œ„ํ—˜

Three static findings are confirmed. Two commands execute a pinned npm package, while the default network path sends request metadata to liarjs.dev. Other detections are markdown or documented local behavior, but optional CDP attachment can access an existing browser session.

2
์Šค์บ”๋œ ํŒŒ์ผ
177
๋ถ„์„๋œ ์ค„ ์ˆ˜
3
๊ฒ€ํ†  ํ•ญ๋ชฉ
0
์˜คํƒ ๋ฌด์‹œ๋จ

ํ™•์ธ๋œ ๋ณด์•ˆ ์šฐ๋ ค ์‚ฌํ•ญ (1)

์ค‘๊ฐ„
Existing Browser Session Access
The optional CDP workflow can drive an existing browser, open tabs, and read page state. Authenticated session content may therefore become accessible.
The skill explicitly states that CDP attachment can open a tab and read page state in the user's session.
๊ธฐ๋Šฅ ๊ฒ€ํ†  ํ•ญ๋ชฉ (3)

์ด ์Šคํ‚ฌ์— ๋Œ€ํ•ด ์˜ˆ์ƒ๋  ์ˆ˜ ์žˆ๋Š” ์‹ค์ œ ๋กœ์ปฌ ๊ธฐ๋Šฅ์ด๋ฏ€๋กœ ๊ฒ€ํ† ๊ฐ€ ํ•„์š”ํ•˜์ง€๋งŒ, ํ™•์ธ๋œ ์•…์˜์  ๋™์ž‘์œผ๋กœ๋Š” ์ง‘๊ณ„๋˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

์ค‘๊ฐ„
Ruby/shell backtick execution
```bash
The Bash block directs the agent to run npx liarjs@0.3, which downloads or executes an external npm package and launches a browser.
์ค‘๊ฐ„
Ruby/shell backtick execution
```bash
The Bash block runs npx liarjs@0.3 and attaches it to a CDP endpoint, executing an external package against a browser session.
์ค‘๊ฐ„
Hardcoded URL
- The network half works by having the browser under test fetch `https://liarjs.dev/api/net.json`,
The default online scan fetches liarjs.dev and exposes request metadata including IP, ASN, headers, HTTP, TLS, and ClientHello characteristics.
๊ฐ์‚ฌ์ž: codex
์ด ๋ณด๊ณ ์„œ ๊ณต์œ  ๋ฐ ์ธ์šฉ

๋ฒ„์ „์ด ์ง€์ •๋œ ํ‰๊ฐ€ ๋ณด๊ณ ์„œ, ์ค‘๋ฆฝ ๋ฐฐ์ง€, ์ž„๋ฒ ๋“œ ์นด๋“œ ๋ฐ ์ธ์šฉ์„ ๊ณต์œ ํ•˜์„ธ์š”. Skillstore๋Š” ๊ทผ๊ฑฐ๋ฅผ ๋ณด๊ณ ํ•˜์ง€๋งŒ ์ด Skill์˜ ์•ˆ์ „ ์—ฌ๋ถ€๋ฅผ ํŒ๋‹จํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

๋ฒ„์ „๋ณ„ ๋ณด๊ณ ์„œ ์—ด๊ธฐ
๋ณด์•ˆ ํ‰๊ฐ€

๋ณด๊ณ ์„œ ๋งํฌ ๋ณต์‚ฌ

https://skillstore.io/skills/liarjsdev-browser-fingerprint-audit/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown ๋ฐฐ์ง€

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

HTML ๋ฐฐ์ง€

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

์ž„๋ฒ ๋“œ ์นด๋“œ

<iframe src="https://skillstore.io/embed/skills/liarjsdev-browser-fingerprint-audit.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
ํ•™์ˆ  ์ธ์šฉ (APA ยท BibTeX ยท CFF)

APA ์ธ์šฉ

liarjsdev. (2026). browser-fingerprint-audit security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/liarjsdev-browser-fingerprint-audit/audits/1

BibTeX ์ธ์šฉ

@techreport{liarjsdev-liarjsdev-browser-fingerprint-audit-2026, author = {liarjsdev}, title = {browser-fingerprint-audit security audit report (audit version 1)}, institution = {Skillstore}, year = {2026}, number = {1}, url = {https://skillstore.io/skills/liarjsdev-browser-fingerprint-audit/audits/1}, 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: "browser-fingerprint-audit security audit report (audit version 1)" version: "unspecified" type: report authors: - name: "liarjsdev" date-released: "2026-08-08" url: "https://skillstore.io/skills/liarjsdev-browser-fingerprint-audit/audits/1" identifiers: - type: other value: "skillstore:liarjsdev-browser-fingerprint-audit:audit:1" description: "Skillstore immutable audit report identifier"

Skillstore ์ ์ˆ˜

์ด ์ ์ˆ˜์˜ ์ด์œ  ์ฆ๊ฑฐ ์‹ ๋ขฐ๋„: ์ค‘๊ฐ„
41
์•„ํ‚คํ…์ฒ˜
85
์œ ์ง€๋ณด์ˆ˜์„ฑ
87
์ฝ˜ํ…์ธ 
65
์ปค๋ฎค๋‹ˆํ‹ฐ
83
์‚ฌ์–‘ ์ค€์ˆ˜

๋งŒ๋“ค ์ˆ˜ ์žˆ๋Š” ๊ฒƒ

Review an automation profile

Identify contradictory browser signals before using a controlled automation environment.

Investigate fingerprint drift

Examine canvas, graphics, font, timezone, and network mismatches after an environment change.

Assess privacy exposure

Run offline checks to understand which local browser characteristics are observable.

์ด ํ”„๋กฌํ”„ํŠธ๋ฅผ ์‚ฌ์šฉํ•ด ๋ณด์„ธ์š”

Run a standard audit
Run the default browser fingerprint audit. Report the score, classification, failed check identifiers, and concise explanations.
Run an offline audit
Run the browser fingerprint audit in offline mode. Explain every failed local check and clearly list the skipped network checks.
Inspect selected signal groups
Audit this browser and focus on canvas, WebGL, WebGPU, fonts, WebRTC, and timezone. Explain contradictions and likely environmental causes.
Audit an existing browser session
Audit my running Chrome through the CDP endpoint I provide. Confirm the endpoint and session access risk before attaching.

๋ชจ๋ฒ” ์‚ฌ๋ก€

  • Use offline mode when external network comparison is unnecessary.
  • Prefer a temporary browser profile unless the existing session is the explicit audit subject.
  • Compare unusual environments with a saved baseline instead of relying only on an absolute score.

ํ”ผํ•˜๊ธฐ

  • Do not attach to an existing browser without explicit approval and endpoint confirmation.
  • Do not treat a high consistency score as proof that every website will trust the browser.
  • Do not send production browsing sessions or sensitive pages to untrusted endpoints.

์ž์ฃผ ๋ฌป๋Š” ์งˆ๋ฌธ

What does the score measure?
It measures consistency among browser signals and the network view. It does not predict every website's detection decision.
Does the standard audit make a network request?
Yes. The online mode contacts liarjs.dev for IP, ASN, HTTP, TLS, header, and ClientHello observations.
Can the audit run without sending data externally?
Yes. Offline mode runs 32 browser-only checks and skips eight network comparison checks.
Does the default scan use my browser profile?
No. It launches a fresh temporary profile and removes that profile after the scan.
Can it scan an already running browser?
Yes, through an approved CDP endpoint. That access can open tabs and read page state in the existing session.
Why can a legitimate environment receive a low score?
Headless mode, datacenter networks, unusual fonts, locale settings, and browser version changes can create real consistency failures.

๊ฐœ๋ฐœ์ž ์„ธ๋ถ€ ์ •๋ณด

์ž‘์„ฑ์ž

liarjsdev

๋ผ์ด์„ ์Šค

MIT

Skillstore ๋ฆฌ๋น„์ „

r1

๋ฒ„์ „ ์•ˆ๋‚ด

์ž‘์„ฑ์ž๊ฐ€ ๋ฒ„์ „์„ ์„ ์–ธํ•˜์ง€ ์•Š์•˜์Šต๋‹ˆ๋‹ค.

์ฐธ์กฐ

17e49992c9f55842f934ecec5719e4745ba7b95a

์œ ์ง€๋ณด์ˆ˜ ์ตœ์‹ ์„ฑ

2026. 8. 9.

์‚ฌ์šฉ๋Ÿ‰

0 ๋‹ค์šด๋กœ๋“œ ยท 0 ์กฐํšŒ

ํŒŒ์ผ ๊ตฌ์กฐ

๐Ÿ“ references/

๐Ÿ“„ checks.md

๐Ÿ“„ SKILL.md