penetration-testing-with-strix

Fail

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download an installation script from https://strix.ai/install and a Python package strix-agent from a public registry.
  • [REMOTE_CODE_EXECUTION]: The skill includes a command that downloads a script and executes it directly in the shell using bash (curl -sSL https://strix.ai/install | bash). This is a high-risk execution pattern as it bypasses local package management and security audits.
  • [COMMAND_EXECUTION]: The skill relies on executing various shell commands, including strix CLI operations, docker management, and setting environment variables for API keys.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill processes data from diverse external sources, including local file paths (-t ./), remote Git repositories, and live URLs/APIs.
  • Boundary markers: There are no specific instructions or delimiters provided to prevent the AI from following malicious instructions potentially embedded in the source code or web applications being scanned.
  • Capability inventory: The skill utilizes a powerful security tool capable of making network requests and interacting with the local file system (via Docker mounts).
  • Sanitization: While the tool aims to validate findings, the initial ingestion of untrusted target data by the AI agents constitutes an attack surface.
Recommendations
  • HIGH: Downloads and executes remote code from: https://strix.ai/install - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 18, 2026, 12:13 AM
Security Audit — agent-trust-hub — penetration-testing-with-strix