The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill follows established patterns for multi-agent coordination and utilizes standard tools for its intended purpose of quality assurance testing.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests data from application UIs which is then processed by AI agents. This is an inherent risk of automated QA testing tools where the application under test could potentially contain malicious content. 1. Ingestion points: Ingests UI data via playwright-cli snapshot (workflows/web-workflow.md), mcp__ios-simulator__ui_describe_all (workflows/mobile-workflow.md), and mcp__mac-mcp-server__get_ui_elements (workflows/macos-workflow.md). 2. Boundary markers: Absent. 3. Capability inventory: The skill has access to Bash, Write, TeamCreate, and SendMessage (SKILL.md). 4. Sanitization: Absent.

qa-team