Skills
Navigate
skills/reason-machines/codex-skills/nanocodex-rust-agent-framework/Gen Agent Trust Hub

nanocodex-rust-agent-framework

Fail

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill documentation instructs users to install a CLI using curl -fsSL https://nanocodex.paradigm.xyz | bash. Piping an external script directly into the bash interpreter allows for arbitrary code execution on the user's system without prior verification, which is a critical security risk.\n- [EXTERNAL_DOWNLOADS]: The installation and update processes involve fetching binaries and scripts from an external domain nanocodex.paradigm.xyz and GitHub PR artifacts. These sources are not recognized as trusted in this context, presenting a supply chain risk.\n- [COMMAND_EXECUTION]: The skill documents commands like nanocodex update --path ./nanocodex and nanocodex update --pr 50 which involve executing code from locally specified paths or unverified remote PR artifacts.\n- [PROMPT_INJECTION]: The FetchUrlTool implementation creates an indirect prompt injection surface by fetching content from arbitrary URLs and passing it directly to the LLM agent without sanitization.\n
  • Ingestion points: FetchUrlTool::execute method in SKILL.md.\n
  • Boundary markers: Absent; the external content is not delimited or accompanied by safety instructions.\n
  • Capability inventory: The agent can perform custom tool execution, access environment variables, and operate within the workspace.\n
  • Sanitization: Absent; raw response text is returned to the agent.
Recommendations
  • HIGH: Downloads and executes remote code from: https://nanocodex.paradigm.xyz - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
CRITICAL
Analyzed
Jul 30, 2026, 04:35 AM
Security Audit — agent-trust-hub — nanocodex-rust-agent-framework