Back to skill

Security audit

anjuke-scraper

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed housing scraper, but it asks to control a logged-in Chrome profile to avoid verification and stores scraped contact data, so it needs careful review before use.

Use only with a dedicated Chrome profile that contains no unrelated logins, keep the CDP port local, confirm that scraping and contact-data collection are allowed for your use case, pin dependencies before installing, sanitize generated spreadsheets before opening, and delete output files containing phone numbers when no longer needed.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Findings (3)

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
scripts/grab_full.py:45
Finding

Overprivileged Access to an Authenticated Browser Context

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/make_excel3.py:58
Finding

Spreadsheet Formula Injection Through Scraped Listing Fields

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Note
Location
requirements.txt:1
Finding

Unbounded Third-Party Dependency Versions

Content
View full analysis
=1.40 openpyxl>=3.1 ``` ### Technical Analysis Both dependencies use open-ended minimum-version constraints. A future installation can therefore resolve to any later release without requiring an explicit project change or security review. This does not establish that either named package is currently malicious. The weakness is that the installed dependency set is not reproducible and lacks integrity verification. Because Playwright has broad authority over an authenticated browser context and openpyxl processes attacker-influenced data, an unexpected or compromised future release would execute with the same privileges as the Skill. No lock file or package hashes are present to verify the exact artifacts installed by `pip`. ### Attack Path 1. A user runs `pip install -r requirements.txt`. 2. The package resolver selects the newest available versions satisfying the open-ended constraints. 3. A later release is compromised, malicious, or contains a newly introduced vulnerability. 4. The unreviewed package is installed because no exact version or artifact hash is enforced. 5. Package code executes during installation or when the Skill imports and uses it. 6. The affected dependency inherits access to local files, generated personal data, network connectivity, and—in Playwright's case—the authenticated browser context. ### Impact Assessment The possible scope is the full privilege level of the Python process running the Skill. This may include access to local output files containing names, phone numbers, addresses, and coordinates, as well as control over the attached Chrome browser context. The current files contain no evidence of dependency confusion, typosquatting, or a known malicious package. Consequently, this is a supply-chain harde ...[truncated 57 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (24)

Natural-Language Policy Violations

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The README instructs users to launch Chrome with remote debugging and reuse a browser profile already logged into Anjuke/58 specifically to avoid extra verification. That is effectively documented bypass of anti-bot and access-control friction using authenticated session state, which increases risk of unauthorized scraping, account misuse, and exposure of sensitive cookies via the debugging interface.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill explicitly instructs use of a user's logged-in Chrome profile and session cookies to access third-party services and preserve verification state. Reusing authenticated browser context for scraping expands access beyond normal unauthenticated browsing and can expose account-linked data or enable actions that rely on the user's session.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill relies on the user's logged-in browser session and cookies but does not present that as a clear upfront warning in the description. Users may not realize that their authenticated state will be reused for scraping and map requests, which materially changes the trust and security implications of running the skill.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The documentation includes specific anti-bot evasion guidance, such as using referers and timing delays to avoid verification challenges. Built-in instructions for defeating anti-automation controls indicate the skill is designed to bypass service protections rather than operate within normal, permitted access patterns.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The README explicitly promotes extraction of broker phone numbers and reuse of authenticated browser state, but provides no meaningful privacy, consent, or data-handling safeguards. In the context of a scraping tool, this enables collection of personal contact data and session-derived access in ways that can violate privacy expectations, platform rules, or local law.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill declares operational capabilities for network access and writing files but does not scope or constrain them via explicit permissions/allowed-tools. That makes the behavior less auditable and increases the risk of unintended data access, uncontrolled outbound requests, and disk persistence of scraped information.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill advertises housing scraping and comparison but does not clearly warn that it extracts broker phone numbers from listing data. Collecting and storing contact information without prominent disclosure increases privacy risk and may cause users to process personal data without informed consent or appropriate handling controls.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill writes multiple intermediate JSON files and a final Excel file containing scraped listing and contact data, but this persistence is not clearly disclosed as a data-handling warning. Undisclosed local storage increases the chance of accidental retention, leakage, or later misuse of collected information.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill describes making authenticated page-context requests to Amap specifically to avoid using an API key. That bypasses normal access controls, obscures usage attribution, and can misuse the user's authenticated session for requests the platform expects to be governed by official API terms.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The script explicitly states it will perform Amap poiInfo requests inside the user's logged-in browser context to avoid using an API key, which means it leverages session cookies and browser trust state for access. That is a security-relevant hidden dependency and can cause the user to unknowingly make authenticated requests under their own account, with possible terms-of-service abuse, privacy exposure, and unintended account activity.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The docstring frames the script as routine coordinate lookup, but also reveals a bypass mechanism: using browser-context fetch with cookies and Referer to avoid API-key usage. That mismatch reduces transparency and can mislead users about the trust and authentication boundaries of the operation, making it easier to hide authenticated scraping behavior.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The code connects to a local Chromium instance over CDP, opens amap.com, and issues fetch requests from that page context, thereby using any active session cookies and browser identity without an explicit user-facing warning. This can silently cause authenticated actions or telemetry under the user's account and is especially risky because the skill is a scraper, where hidden use of session state materially changes the security posture.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The script explicitly extracts mobile phone numbers from page content and stores them in detail_results.json without any consent checks, minimization, masking, retention controls, or user warning. In this skill’s context, scraping rental listings and broker contact details makes the privacy risk more acute because it operationalizes bulk collection of personal contact data that could be reused for spam, profiling, or unauthorized outreach.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This file contains natural-language strings such as the module docstring and console messages entirely in Chinese, which effectively forces a specific language for users. Under the policy, language constraints should either be optional for the user or clearly documented as justified and locale-specific.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script explicitly extracts phone numbers from listing HTML and URL-derived IDs, then stores them in structured records. In the context of a scraper skill for housing listings, this creates a privacy/compliance risk because personally identifiable contact data is collected and normalized without any consent check, minimization control, or notice mechanism.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The output file final_records.json persists structured listing data that includes extracted phone numbers and contact names. Persisting PII to disk increases exposure through accidental sharing, reuse, or compromise of the output directory, especially since the skill is designed for scalable repeated scraping and comparison workflows.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
94% confidence
Finding

The dependency specification uses a lower-bound only constraint for Playwright, which makes builds non-reproducible and allows unreviewed future versions to be installed. This increases supply-chain risk and can introduce breaking or vulnerable releases without the maintainer noticing.

Content

Scanner excerpt · requirements.txt (reported line 1)May include surrounding context.

text
playwright>=1.40
openpyxl>=3.1

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
97% confidence
Finding

The openpyxl dependency is also specified with only a minimum version, so installations may resolve to different versions over time. For a scraping skill that generates Excel files, this creates avoidable supply-chain and reliability risk because a future vulnerable or incompatible release could be pulled in automatically.

Content

Scanner excerpt · requirements.txt (reported line 2)May include surrounding context.

text
playwright>=1.40
openpyxl>=3.1

Unverifiable Dependency: openpyxl has 2 known advisory(ies) (CVE-2017-5992 (Improper Restriction of XML External Entity Reference in Openpyxl); CVE-2017-5992 (Openpyxl 2.4.1 resolves external entities by default, which allows remote attack)), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
92% confidence
Finding

The manifest does not pin openpyxl, so it is not possible to verify whether deployment will use a version affected by known XXE-related advisories. In this skill's context, openpyxl is used for Excel generation rather than clearly parsing untrusted XLSX input, which lowers direct exploitability, but the unresolved version risk remains real.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The file's user-facing docstring and messages are written entirely in Chinese, including usage guidance, with no indication that another language is supported or that Chinese is required for a region-specific reason. This can violate a language/locale policy when a skill mandates one language without opt-in or justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The top-level docstring and subsequent user-facing messages are written in Chinese, and the script assumes Chinese-language page structure and field labels. This enforces a specific language/locale experience without any user opt-in or documented locale constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The top-level docstring presents the script description and usage entirely in Chinese, which imposes a specific language on users. There is no indication of language choice, bilingual support, or justification for the locale restriction.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The script persists collected listing text and HTML into cards_full.json and all_pages.html, which affects local user data/storage. While the module docstring mentions that data will be saved, the write operations themselves have no inline warning, confirmation, or explicit disclosure about the specific files and contents being created.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The natural-language docstring and user-facing messages are entirely in Chinese, which imposes a specific language/locale on users without opt-in or explanation. Under the policy, locale-specific behavior should either offer a language choice or clearly document that the skill is intentionally region/language-specific.

Content

No source excerpt is available for this finding.