lazy-agent-loader

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMSAFE
Full Analysis
  • Dynamic Execution (MEDIUM): The skill utilizes dynamic loading from computed paths. It constructs file paths as agents/[agent-id].md at runtime based on the selection logic from the agent_index. This matches the criteria for medium severity dynamic loading.
  • Indirect Prompt Injection Surface (LOW): The skill possesses a data ingestion surface that processes external content (agent definitions in agents/).
  • Ingestion points: The skill reads files matching agents/*.md.
  • Boundary markers: Absent. There are no instructions to the agent to treat the loaded content as data rather than instructions.
  • Capability inventory: The skill is authorized to use Read and Glob tools.
  • Sanitization: None detected. The content of the loaded files is interpolated directly into the session context.
  • Tool Usage (SAFE): The skill is limited to Read and Glob tools, which are appropriate for its stated purpose of managing and loading local agent definitions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:34 PM