hermes-webui-agent
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to clone the hermes-webui repository from GitHub and pull Docker images from GitHub Container Registry (ghcr.io).\n- [REMOTE_CODE_EXECUTION]: Installation instructions guide the user to execute a bootstrap Python script (bootstrap.py) and shell scripts (start.sh, ctl.sh) after cloning the external repository.\n- [COMMAND_EXECUTION]: Python code snippets demonstrate the use of subprocess.run to manage the lifecycle of the web server and interact with the Hermes Agent SDK.\n- [PERSISTENCE]: The skill describes multiple methods for maintaining the agent's operation, including shell-based daemons, cron-scheduled jobs for periodic tasks, and Windows VBS scripts for startup execution.\n- [INDIRECT_PROMPT_INJECTION]: The skill documents how the agent ingests and processes data from external sources such as Telegram, Discord, and Slack.\n
- Ingestion points: Messaging platform webhooks and the WebUI chat interface.\n
- Boundary markers: None explicitly defined in the provided instruction snippets.\n
- Capability inventory: The skill uses subprocess.run and file system writes (~/.hermes) to manage state and execute tasks.\n
- Sanitization: Not explicitly covered in the provided configuration examples.
Audit Metadata