bb-subagents

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides recipes for executing local CLI commands using the bb tool. These commands are used for listing projects, providers, models, and spawning new threads. This is the primary intended purpose of the skill.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a mechanism for passing instructions to sub-agents via the --prompt flag.
  • Ingestion points: The agent ingests user requirements to define the sub-agent's objective and task (SKILL.md).
  • Boundary markers: The instructions mandate the use of shell heredocs (cat <<'EOF') to prevent command injection into the shell and suggest a structured format for the prompt content (Objective, Constraints, Deliverable, etc.).
  • Capability inventory: The skill facilitates spawning new worker threads using bb thread spawn, which supports various permission levels including --permission-mode full.
  • Sanitization: The use of heredocs ensures that user-provided text does not break the shell command structure, although the content of the prompt remains a surface for indirect instructions to the sub-agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 12:12 PM
Security Audit — agent-trust-hub — bb-subagents