orchestrator
Pass
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is focused on task orchestration and uses internal platform tools (codex_app.*) for thread lifecycle management. No external network requests or suspicious file system operations were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill implements a routing mechanism that forwards user input to child threads. This is the primary function of the skill, though it technically creates a surface for indirect prompt injection.
- Ingestion points: User requests and task descriptions provided to the orchestrator thread (SKILL.md).
- Boundary markers: The skill uses a structured Markdown template (Child Prompt Shape) to wrap instructions, providing context delimiters.
- Capability inventory: Uses thread management tools like codex_app.send_message_to_thread to move data between contexts (SKILL.md).
- Sanitization: No explicit sanitization or filtering of the forwarded content is mentioned; the skill relies on the prompt template to maintain task boundaries.
Audit Metadata