using-designpowers

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill uses extremely assertive and mandatory language to override default agent behavior and hijack the response loop.
  • Evidence in SKILL.md: "Use when starting any conversation — establishes how to find and use design skills, requiring Skill tool invocation before ANY response including clarifying questions"
  • Evidence in SKILL.md: "IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT."
  • Evidence in SKILL.md: "Before responding to ANY message — including clarifying questions — check whether a Designpowers skill applies."
  • [PROMPT_INJECTION]: It mandates a multi-step "Welcome Sequence" that must be executed "before doing anything else" in a session, regardless of the initial user prompt.
  • [PROMPT_INJECTION]: Indirect prompt injection surface via untrusted data ingestion.
  • Ingestion points: User response to AskUserQuestion ("What are we designing?") and agent-generated "handoff babble".
  • Boundary markers: Absent for user input.
  • Capability inventory: Invocation of numerous other tools and skills (e.g., design-discovery, ui-composition), and writing to a shared state file (design-state.md).
  • Sanitization: Absent; the skill does not specify any validation or filtering of user-provided project descriptions before they are passed to subsequent agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 05:42 PM