hermes-edu-skills

Warn

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install an external package hermes-edu-skills from the NPM registry. This package originates from an unverified repository (github.com/zhongweiv/hermes-edu-skills) not associated with recognized trusted vendors.
  • [REMOTE_CODE_EXECUTION]: The documentation encourages the use of npx hermes-edu-skills, which downloads and executes remote code at runtime. This allows the execution of code directly from the registry without prior local verification.
  • [COMMAND_EXECUTION]: Setting up the skill involves running several shell commands, including global package installation (npm install -g) and automated configuration scripts that modify system files.
  • [DATA_EXFILTRATION]: The skill's installation logic modifies the configuration file ~/.hermes/config.yaml. Automated modification of configuration files by third-party tools can be leveraged for persistence or to manipulate agent behavior if the package is compromised.
  • [PROMPT_INJECTION]: The skill processes natural language queries and user-uploaded images to route to specific educational functions. This creates a surface for indirect prompt injection, as the skill lacks defined boundary markers or instructions to sanitize potentially malicious content embedded in the processed educational data.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 14, 2026, 08:47 AM
Security Audit — agent-trust-hub — hermes-edu-skills