task-orchestration

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The orchestration workflows in references/spawn.md and references/task.md utilize the bash tool for system-level operations and coordination. This enables the agent to execute commands within the environment to manage complex tasks during orchestration.
  • [PROMPT_INJECTION]: Surface for Indirect Prompt Injection.
  • Ingestion points: The skill accepts user-provided topics in references/brainstorm.md (topic/idea) and task descriptions in references/task.md (action/target) and references/spawn.md (complex-task).
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present for the processed data.
  • Capability inventory: The skill utilizes bash, Task (for spawning subagents), and file manipulation tools (Read, Write, Edit, TodoWrite).
  • Sanitization: No specific sanitization or validation logic for the input strings is described.
  • [SAFE]: The skill incorporates robust software engineering principles for distributed systems, including the Saga pattern for multi-step transactions, circuit breakers for fault tolerance, and idempotency for reliable retries.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 05:14 PM