📦

История аудитов

playwright-cli - 4 аудиты

Сравнение версий

Изменения возможностей и находок между проверенными версиями, сначала новые.

Версия	Дата	Результат	Пункты проверки	Изменение к предыдущей
v4 Последняя	5 июл. 2026 г., 23:10	10 подтверждено	0	Возможности не изменились
v3	5 июл. 2026 г., 23:10	10 подтверждено	0	Содержит скрипты
v2	30 июн. 2026 г., 08:07	4 подтверждено	0	Содержит скрипты
v1	24 мар. 2026 г., 08:23	Подтверждённых находок нет	1	Базовая

Версия аудита 4 Последняя

5 июл. 2026 г., 23:10

Most static findings are false positives from Markdown examples, placeholder URLs, and documented playwright-cli commands. Confirmed risks are limited to browser-session data operations and documented advanced capabilities such as run-code, tracing, and destructive session cleanup.

Просканировано файлов

1,149

Проанализировано строк

Пункты проверки

Ложные срабатывания проигнорированы

Подтверждённые проблемы безопасности (10)

Средний

SKILL.md:147-152 references/running-code.md:1-11

Arbitrary Playwright Code Execution

The skill documents run-code as a way to execute arbitrary Playwright code. This is powerful and expected, but it can read page data, grant permissions, or alter browser state on sensitive sites.

The documentation explicitly describes run-code for arbitrary Playwright code and shows direct browser context operations. The risk depends on user intent, but the capability is clear.

Средний

SKILL.md:111-118 references/storage-state.md:31-56 references/storage-state.md:269-275

Browser Session Data Access

The skill supports reading, writing, deleting, saving, and restoring cookies and browser storage. These operations can expose or modify authentication state if used on real accounts.

The cited lines show cookie commands, stored cookies and localStorage values, and security notes about auth tokens. This confirms the session data exposure risk.

Средний

references/tracing.md:34-40 references/tracing.md:87-100

Trace Artifacts Can Capture Sensitive Data

Tracing captures request headers, bodies, DOM snapshots, screenshots, and console messages. The example flow enters payment-like values, so traces can store sensitive test or user data.

The documentation states traces include request and response bodies and shows a checkout capture flow. That directly supports the sensitive artifact risk.

Низкий

references/storage-state.md:82-83

System reconnaissance

playwright-cli cookie-get session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

references/storage-state.md:101-102

System reconnaissance

playwright-cli cookie-delete session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:114-115

System reconnaissance

playwright-cli cookie-get session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:115

System reconnaissance

playwright-cli cookie-set session_id abc123

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:116

System reconnaissance

playwright-cli cookie-set session_id abc123 --domain=example.com --httpOnly --secure

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:117-118

System reconnaissance

playwright-cli cookie-delete session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:176-180 SKILL.md:201-214

Destructive Browser Session Controls

The skill documents commands that delete session data and forcefully kill browser processes. These commands are legitimate, but they can disrupt active work or remove useful state.

The cited commands explicitly delete user data, close browsers, and kill browser processes. The impact is operational disruption rather than malicious behavior.

Факторы риска

🌐 Доступ к сети (34)

⚙️ Внешние команды (39)

📁 Доступ к файловой системе (1)

references/tracing.md:107

Обнаруженные паттерны

System reconnaissanceSystem reconnaissanceSystem reconnaissanceSystem reconnaissanceSystem reconnaissanceSystem reconnaissance

Аудитор:: codex

Версия аудита 3

5 июл. 2026 г., 23:10

Просканировано файлов

1,149

Проанализировано строк

Пункты проверки

Ложные срабатывания проигнорированы

Подтверждённые проблемы безопасности (10)

Средний

SKILL.md:147-152 references/running-code.md:1-11

Arbitrary Playwright Code Execution

The skill documents run-code as a way to execute arbitrary Playwright code. This is powerful and expected, but it can read page data, grant permissions, or alter browser state on sensitive sites.

The documentation explicitly describes run-code for arbitrary Playwright code and shows direct browser context operations. The risk depends on user intent, but the capability is clear.

Средний

SKILL.md:111-118 references/storage-state.md:31-56 references/storage-state.md:269-275

Browser Session Data Access

The skill supports reading, writing, deleting, saving, and restoring cookies and browser storage. These operations can expose or modify authentication state if used on real accounts.

The cited lines show cookie commands, stored cookies and localStorage values, and security notes about auth tokens. This confirms the session data exposure risk.

Средний

references/tracing.md:34-40 references/tracing.md:87-100

Trace Artifacts Can Capture Sensitive Data

Tracing captures request headers, bodies, DOM snapshots, screenshots, and console messages. The example flow enters payment-like values, so traces can store sensitive test or user data.

The documentation states traces include request and response bodies and shows a checkout capture flow. That directly supports the sensitive artifact risk.

Низкий

references/storage-state.md:82-83

System reconnaissance

playwright-cli cookie-get session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

references/storage-state.md:101-102

System reconnaissance

playwright-cli cookie-delete session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:114-115

System reconnaissance

playwright-cli cookie-get session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:115

System reconnaissance

playwright-cli cookie-set session_id abc123

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:116

System reconnaissance

playwright-cli cookie-set session_id abc123 --domain=example.com --httpOnly --secure

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:117-118

System reconnaissance

playwright-cli cookie-delete session_id

The command can read, set, or delete browser cookies such as session_id, which may represent authentication state. This is documented functionality, but it is a real session data risk.

Низкий

SKILL.md:176-180 SKILL.md:201-214

Destructive Browser Session Controls

The skill documents commands that delete session data and forcefully kill browser processes. These commands are legitimate, but they can disrupt active work or remove useful state.

The cited commands explicitly delete user data, close browsers, and kill browser processes. The impact is operational disruption rather than malicious behavior.

Факторы риска

🌐 Доступ к сети (34)

⚙️ Внешние команды (39)

📁 Доступ к файловой системе (1)

references/tracing.md:107

Обнаруженные паттерны

System reconnaissanceSystem reconnaissanceSystem reconnaissanceSystem reconnaissanceSystem reconnaissanceSystem reconnaissance

Аудитор:: codex

Версия аудита 2

30 июн. 2026 г., 08:07

The static command findings are mostly false positives from markdown command examples, not Ruby backtick execution. The skill is still medium risk because it intentionally allows browser automation, arbitrary Playwright page code, network navigation, storage-state files, persistent profiles, and trace artifacts that can expose sensitive session data. No prompt injection attempt or confirmed malicious exfiltration path was found.

Просканировано файлов

1,149

Проанализировано строк

Пункты проверки

Ложные срабатывания проигнорированы

Подтверждённые проблемы безопасности (4)

Средний

SKILL.md:104-132 references/storage-state.md:7-27 references/storage-state.md:231-247 references/storage-state.md:269-275

Browser Storage State Can Expose Sessions

The skill documents saving, loading, listing, and editing cookies, localStorage, sessionStorage, and authentication state files. This is legitimate for testing, but saved state can contain session tokens and must be protected.

The lines explicitly show state-save, state-load, cookie commands, and security notes about auth tokens. The context is legitimate automation, so the risk is exposure through misuse rather than confirmed malicious intent.

Средний

SKILL.md:47-48 SKILL.md:147-151 references/running-code.md:1-11 references/running-code.md:173-185

Arbitrary Playwright Code Execution in Browser Context

The skill supports run-code and eval commands that execute user supplied Playwright or JavaScript snippets in the active browser session. This enables advanced testing, but it can read page content, grant browser permissions, and manipulate storage.

The commands are directly documented and the examples show page evaluation and permission changes. The capability is expected for a Playwright CLI skill, but it increases impact if misused.

Средний

references/tracing.md:1-3 references/tracing.md:34-43 references/tracing.md:87-100 references/video-recording.md:1-18

Trace and Video Artifacts May Capture Sensitive Data

Tracing records DOM snapshots, screenshots, network requests, headers, bodies, and console logs. Video recording and traces can preserve secrets typed during automated flows.

The reference states that traces capture network bodies, DOM snapshots, screenshots, and checkout form inputs. This is strong evidence of sensitive artifact risk without evidence of automatic exfiltration.

Средний

SKILL.md:168-179 SKILL.md:201-214

Persistent Browser Profiles Retain Local Data

The skill supports persistent sessions and custom profile directories. Persistent profiles can keep cookies, local storage, and other browser data beyond the immediate automation task.

The documented commands explicitly create persistent sessions and provide cleanup commands. The risk is manageable, but users must understand retention and delete data after sensitive workflows.

Статические ложные срабатывания проигнорированы (3)

Эти статические совпадения были отклонены семантической проверкой или совпадали только со схемными токенами, поэтому они показываются для прозрачности, но не влияют на оценку качества.

Низкий

SKILL.md:11-23 references/request-mocking.md:7-26 references/running-code.md:7-12 references/test-generation.md:12-32

Static External Command Matches Are Mostly Documentation Examples

The many Ruby or shell backtick detections occur in fenced markdown blocks that show playwright-cli usage. They are not Ruby backtick execution in source code, although the skill intentionally allows the playwright-cli command.

The cited locations are markdown command examples, not executable Ruby code. The allowed Bash wrapper is real, but the specific static Ruby backtick label is a false positive.

Низкий

SKILL.md:15 SKILL.md:33-34 references/running-code.md:210-226 references/tracing.md:118-123

Hardcoded URLs Are Example Targets

The hardcoded URL findings use documentation domains such as example.com and playwright.dev. I did not find evidence that these URLs receive secrets or hidden telemetry.

The URL examples are clearly part of user-facing command documentation and test workflows. No evidence found of a fixed third-party collection endpoint.

Низкий

SKILL.md:1-5

Prompt Injection Search Found No Evidence

No reviewed file contained suspicious override language, fake system instructions, pre-approval claims, or requests to skip analysis.

Targeted review of the skill and references found no prompt injection phrases. The confidence is high, but this finding is limited to the files included in the audit.

Факторы риска

⚙️ Внешние команды (4)

SKILL.md:4 SKILL.md:11-23 SKILL.md:147-155 SKILL.md:219-223

⚡ Содержит скрипты (4)

SKILL.md:47-48 references/running-code.md:1-11 references/storage-state.md:112-120 references/storage-state.md:165-172

🌐 Доступ к сети (4)

SKILL.md:135-142 references/request-mocking.md:1-18 references/request-mocking.md:56-66 references/tracing.md:34-43

📁 Доступ к файловой системе (4)

SKILL.md:84-90 SKILL.md:168-179 references/running-code.md:120-131 references/tracing.md:20-49

Обнаруженные паттерны

Bash Tool Boundary Allows Playwright CLI Executionrun-code Provides Custom Automation LogicAuthentication Artifacts Are Written to Files

Аудитор:: codex

Версия аудита 1

24 мар. 2026 г., 08:23

Static analysis flagged 289 patterns (187 external_commands, 51 network, 1 filesystem) but all are FALSE POSITIVES. The 'Ruby/shell backtick execution' patterns are markdown documentation examples, not actual code execution. Network findings are URLs in documentation for browser navigation examples. Browser credential/storage access is intentional functionality for testing authenticated workflows. This is a legitimate Microsoft Playwright CLI wrapper skill with appropriate tool scoping via allowed-tools directive.

Просканировано файлов

1,319

Проанализировано строк

Пункты проверки

Ложные срабатывания проигнорированы

Пункты проверки возможностей (1)

Это реальные локальные возможности, которые могут ожидаться для этого навыка, поэтому они требуют проверки, но не считаются подтверждённым вредоносным поведением.

Низкий

SKILL.md:4

External Command Execution Capability

Skill can execute arbitrary playwright-cli commands via Bash. This is by design but requires user trust in command outputs.

Факторы риска

⚙️ Внешние команды (1)

SKILL.md:4

🌐 Доступ к сети (1)

SKILL.md:15

📁 Доступ к файловой системе (1)

SKILL.md:88-90

Аудитор:: claude

← Назад к навыку