paw-cra-campaign-orchestration
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Agent tool to orchestrate tasks across specialized sub-agents including
paw-cra-agent-designer,paw-cra-agent-video-producer,paw-cra-agent-strategist, andpaw-cra-quality-control. This multi-agent coordination is the primary intended function of the orchestration engine and is performed within the scope of the project's creative suite directory.\n- [PROMPT_INJECTION]: As an orchestrator, the skill processes campaign briefs which are untrusted external inputs. This ingestion presents a surface for indirect prompt injection, although the structured processing pipeline mitigates the risk.\n - Ingestion points: Campaign brief files (e.g.,
brief.md) parsed in Stage 01.\n - Boundary markers: Deliverables are extracted and classified into structured lists before being passed to sub-agents with focused instruction sets rather than raw brief interpolation.\n
- Capability inventory: The skill possesses file read/write capabilities within the
.pawbytes/project directory and invokes sub-agents via tool calls.\n - Sanitization: Stage 01 performs deliverable classification and dependency analysis, effectively filtering raw input into structured production tasks.
Audit Metadata