autogpt-agents

Pass

Audited by Gen Agent Trust Hub on Oct 1, 2026

Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [DYNAMIC_EXECUTION]: The platform facilitates the creation and execution of custom logic through its 'Forge' development toolkit and the implementation of custom blocks in Python. This capability allows for runtime extension of agent functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill documents how to ingest data from external sources, including webhooks and integrations with services like GitHub and Discord. This data is processed by LLM blocks, creating a surface where malicious instructions embedded in external content could potentially influence agent behavior.
  • [COMMAND_EXECUTION]: The instructions include the use of CLI tools for agent creation, setup, and performance benchmarking, as well as Docker commands for service management.
  • [EXTERNAL_DOWNLOADS]: The skill directs users to clone the official AutoGPT source code from the Significant-Gravitas GitHub repository as part of the standard installation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 1, 2026, 07:50 AM
Security Audit — agent-trust-hub — autogpt-agents