grace-setup-subagents

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE]: The skill scans for agent configuration files in the current project, user-level global config directories, and neighboring projects to determine the appropriate file format for the shell environment. While this involves reading local configuration metadata, it is a standard part of the scaffolding and environment detection process.
  • [INDIRECT_PROMPT_INJECTION]: The skill uses content from local configuration files as templates to infer metadata fields, model names, and tool permissions for the newly generated agents. This creates a surface where a malicious local file could theoretically influence the structure of the scaffolded agents, though the risk is low and inherent to the tool's primary purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 08:08 AM