ecosystem-database
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/install_all.pyutilizes thesubprocessmodule to execute theskillsCLI tool on the host system. This is used to install a static list of ten sub-skills related to the database ecosystem. The script does not accept external inputs for these commands, reducing the risk of command injection. - [PROMPT_INJECTION]: The
SKILL.mdfile contains directives that instruct the AI agent to prioritize specific workflows and avoid 'hallucinating' steps. While these are strong behavioral constraints, they are intended to maintain operational reliability within the database domain rather than overriding the agent's safety or ethical guidelines.
Audit Metadata