Back to skill

Security audit

salesman-recommender

Security checks for vulnerabilities and agentic risk

Overview

This skill can perform the stated creator export task, but it also automates account login, reuses browser sessions, intercepts API traffic, and can save sensitive debug files without enough user control.

Install only if you are comfortable letting this skill automate a logged-in Buyin/Jinritemai business account. Use a dedicated account/profile if possible, review and delete any Desktop debug files, verify the selected account before relying on results, and treat generated Excel files as untrusted because creator-controlled fields are not formula-sanitized.

Vulnerability Patterns
  • 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
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (3)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/daren_search.py:1909
Finding

Spreadsheet Formula Injection in Excel Export

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/daren_search.py:1158
Finding

Sensitive Login and Account Data Persisted in Desktop Debug Artifacts

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
requirements.txt:1
Finding

Unpinned and Unverified Third-Party Dependencies

Content
View full analysis
=4.15.0 pandas>=2.0.0 openpyxl>=3.1.0 ``` ### Technical Analysis All dependencies use open-ended lower-bound constraints. The project does not provide exact versions, package hashes, or a lock file. Consequently, two installations performed at different times may resolve to different dependency versions. Any future release satisfying these constraints can be installed without project-level review. This increases exposure to: - Compromised upstream releases. - Newly introduced vulnerabilities. - Unexpected or incompatible behavior. - Transitive dependency changes. - Non-reproducible security testing. These packages operate with the same privileges as the script. Selenium controls an authenticated browser, while pandas and `openpyxl` process and write local data. A compromised package would therefore execute with access to the user's files, environment, and browser automation process. ### Attack Path 1. The user or deployment process installs `requirements.txt`. 2. The package resolver selects the newest releases satisfying the lower bounds. 3. A selected direct or transitive dependency contains malicious code or a newly introduced vulnerability. 4. Installation or import activates the affected package code. 5. The compromised component runs with the Python process's privileges and can potentially access local files, browser state, credentials in memory, and exported data. This is a supply-chain exposure rather than evidence that the currently named packages are malicious. ### Impact Assessment The maximum impact depends on the behavior of a compromised dependency and the privileges of the user running the Skill. Because dependencies execute in-process, potential scope includes: - Reading or altering files accessible to the user ...[truncated 338 chars]
Remediation
View remediation
--hash=sha256: pandas== --hash=sha256: openpyxl== --hash=sha256: ``` The actual versions and hashes should be generated from reviewed release artifacts rather than copied from an unverified source. ]]>
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 (21)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The documented purpose is simple filtering/export, but the skill also handles credentials, persists sessions via the default Chrome profile, intercepts API/XHR data, manipulates browser storage, and saves debugging artifacts locally. This mismatch is dangerous because users may consent to a narrow data-export task without realizing the skill can access account sessions, collect sensitive tokens or page contents, and leave credential or session material on disk.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill performs network access and writes files but does not declare any explicit tool scope or permission boundary. This makes the capability set opaque to reviewers and increases the chance of unintended data access, silent file creation, or privilege creep when the skill is executed in an agent environment.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill name includes a CN locale marker and the entire user-facing description is written only in Chinese, with no indication that users may choose another language or that the skill is restricted to a China-specific audience for compliance reasons. This can violate language/locale policy when a skill implicitly forces one language without opt-in.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script persists a browser profile under the user's home directory and reuses cookies/session state across runs while also automating login. That expands the skill beyond simple filtering/export and creates credential/session exposure risk if the local machine, profile directory, or subsequent runs are compromised.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The browser configuration disables automation indicators and related Selenium flags to evade bot detection. Anti-detection behavior is unnecessary for the stated export task and is a strong signal of attempts to circumvent platform defenses, which can facilitate unauthorized automation at scale.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The code injects JavaScript/CDP hooks to capture full XHR request/response bodies for the internal seekAuthor API, including data not limited to what the visible UI intentionally exposes. Intercepting and retaining raw network responses increases the chance of collecting sensitive fields, bypassing intended access boundaries, and violating user expectations or platform controls.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The script captures and stores full XHR responses without clear user-facing notice, and those responses may contain sensitive account, creator, or internal platform data beyond the final export set. Silent network capture materially increases privacy and compliance risk because users may not realize raw API traffic is being collected in memory and processed.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The script accepts email and password, automates login, and may persist session state, but it provides no strong safeguards for credential handling. In this context, credentials could be exposed through logs, browser profile persistence, screenshots/page dumps during failures, or misuse of the automated session.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

After login, the script auto-selects an account and even defaults to the second visible account when multiple accounts are present. This can cause actions and data collection to occur under the wrong identity or tenant, leading to unintended access, privacy issues, and cross-account data exposure.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Saving full page source to local disk for debugging can capture tokens, account identifiers, personal data, or other sensitive HTML/embedded state. Because this happens on error paths and writes to predictable locations like the Desktop, it increases accidental local disclosure risk.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The export function writes structured creator data, including IDs, handles, city, follower counts, and other derived metrics, to a local spreadsheet. The tool announces success after writing, but it does not clearly disclose beforehand that collected third-party data will be stored on disk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill name includes the locale suffix "CN", which indicates a China-specific language or locale constraint. In this file there is no accompanying natural-language justification or indication that the user can opt into that locale, so it may violate the policy against forcing a specific language/locale by default.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
95% confidence
Finding

The dependency uses a lower-bound version specifier instead of pinning to an exact version, which makes builds non-reproducible and can allow unexpected or vulnerable releases to be installed later. In an automation skill that depends on browser-driving libraries like Selenium, this increases supply-chain and stability risk even if it is not directly exploitable on its own.

Content

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

text
selenium>=4.15.0
pandas>=2.0.0
openpyxl>=3.1.0

Unverifiable Dependency: selenium has 2 known advisory(ies) (CVE-2022-28108 (Selenium Server (Grid) before 4 allows CSRF because it permits non-JSON content ); CVE-2023-5590 (NULL Pointer Dereference in GitHub repository seleniumhq/selenium prior to 4.14.)), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
40% confidence
Finding

Dependency has known vulnerabilities (CVEs). Using packages with unpatched security flaws exposes the environment to known exploits.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
94% confidence
Finding

Using pandas>=2.0.0 allows any future compatible version to be installed, which weakens build determinism and can introduce vulnerable or behavior-changing releases without review. This is primarily a supply-chain hygiene issue rather than an immediate exploit, but it is still a real security weakness.

Content

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

text
selenium>=4.15.0
pandas>=2.0.0
openpyxl>=3.1.0

Unverifiable Dependency: pandas has 1 known advisory(ies) (CVE-2020-13091 (** DISPUTED ** pandas through 1.0.3 can unserialize and execute commands from an)), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
40% confidence
Finding

Dependency has known vulnerabilities (CVEs). Using packages with unpatched security flaws exposes the environment to known exploits.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
94% confidence
Finding

The openpyxl dependency is unpinned, so package resolution may pull different versions over time, including versions with newly disclosed flaws or incompatible behavior. Because this skill writes Excel files, the library is directly in the execution path, making version control important for security and reliability.

Content

Scanner excerpt · requirements.txt (reported line 3)May include surrounding context.

text
selenium>=4.15.0
pandas>=2.0.0
openpyxl>=3.1.0

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
40% confidence
Finding

Dependency has known vulnerabilities (CVEs). Using packages with unpatched security flaws exposes the environment to known exploits.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

Natural-language strings throughout the file assume a Chinese UI and Chinese-language user environment, starting from the module description and continuing through selectors and prompts. The skill does not offer any opt-in or alternative locale handling, which can violate language/locale choice policy when not explicitly justified.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

This documentation asserts that the login form is not inside an iframe, yet the calling login flow and input-finding logic include iframe enumeration and interaction paths. That is an active contradiction between the documented page model and actual supported behavior.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The docstring describes a main-page-only structure, but later in the same function the code explicitly iterates over iframes to locate login inputs. This is a direct contradiction, not merely missing detail.

Content

No source excerpt is available for this finding.