Back to skill

Security audit

Legionspace Version

Security checks for vulnerabilities and agentic risk

Overview

The skill does the stated version-checking task, but it uses unusually unsafe network and browser settings for that task.

Review before installing. The skill is not clearly malicious, but it should be run only in a restricted environment if used at all, because it may install new packages, bypass configured proxies, accept invalid HTTPS certificates, and run remote web pages in a browser without sandboxing. Prefer a revised version that pins dependencies, respects proxy settings, keeps TLS validation enabled, and avoids --no-sandbox.

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)

T08 · Insecure Dependencies

Error
Location
SKILL.md:24
Finding

Unpinned Runtime Installation of Third-Party Dependencies

Content
View full analysis
/dev/null || python -m pip install requests python -c "from playwright.sync_api import sync_playwright" 2>/dev/null || { python -m pip install playwright && python -m playwright install chromium; } ``` ### Technical Analysis The Skill directs the agent to install the latest available releases of `requests`, Playwright, and Playwright's Chromium artifact when they are absent. No exact versions, package hashes, lock file, trusted repository configuration, or artifact integrity checks are specified. Consequently, the effective code executed by the Skill can change after the Skill itself has been audited. Python packages may execute code during installation or when imported, while the downloaded browser is a native executable. A compromised upstream release, dependency-confusion condition, malicious package-index configuration, or compromised artifact distribution channel could introduce attacker-controlled code. ### Attack Path 1. An attacker compromises an upstream package release, transitive dependency, configured Python package index, or Playwright browser artifact. 2. The Skill runs in an environment where one or more required components are absent. 3. The dependency checks fail, causing the unpinned installation commands to execute. 4. The package manager resolves the mutable latest package or artifact from the affected source. 5. Malicious installation, import-time, or executable code runs with the same operating-system privileges as the agent. 6. The malicious component can access files, environment variables, network resources, and credentials available to that account. ### Impact Assessment Successful exploitation permits arbitrary code execution under the agent's current account. The affected scope includes the project works ...[truncated 267 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:115
Finding

Chromium Sandbox and TLS Certificate Validation Disabled

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:39
Finding

Mandatory Proxy and Network Security Controls Are Bypassed

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill title and operating instructions are presented in Chinese and describe the task exclusively in that locale context, without stating that users may choose another language. Under the policy, forcing a specific language or locale without user opt-in is a natural-language policy violation unless the constraint is clearly documented and justified as region-specific.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The script globally disables proxy settings and forces direct outbound connections, bypassing environment-based network controls that may be required for monitoring, DLP, egress policy, or enterprise security inspection. For a version-checking skill, this exceeds necessity and can undermine security boundaries in managed environments.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

Launching Chromium with --no-sandbox disables an important browser security isolation mechanism. Even though the skill only visits app-store pages, it loads remote untrusted web content in a headless browser, so removing sandboxing unnecessarily increases the blast radius of a browser compromise.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest describes checking LegionSpace versions across app stores, which clearly justifies network access and scraping, but it does not mention creating local artifacts. The code additionally persists results to ./LegionSpace_versions.txt, which is beyond a pure check/query behavior described at the manifest level.

Content

No source excerpt is available for this finding.