hermes-agent-framework
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download the Hermes Agent framework from the official Nous Research GitHub repository and install the
hermes-agentpackage using standard Python package management tools. - [DYNAMIC_EXECUTION]: The framework utilizes dynamic execution to implement custom agent 'Skills', which are stored as Python code strings. The documentation includes a safety configuration section showing how to restrict shell commands and specific operations to maintain security.
- [INDIRECT_PROMPT_INJECTION]: By enabling tools like web scraping and API searching, the framework creates an ingestion surface for external data that could contain indirect prompt injections. To address this, the documentation demonstrates the use of a
ConstraintConfigto set safety boundaries and require approval for sensitive operations.
Audit Metadata