product-launch-war-room

Pass

Audited by Gen Agent Trust Hub on Jun 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes content from external web sources and local files to ground its simulations.
  • Ingestion points: The skill uses WebSearch and WebFetch to gather market context and competitor information, and the Read tool to load persona data from the personas/ directory.
  • Boundary markers: There are no explicit delimiters or boundary markers (such as XML tags or triple backticks with 'ignore' instructions) defined to isolate untrusted external content from the core logic.
  • Capability inventory: The agent is granted access to high-impact tools including Bash, Write, and the ability to delegate tasks to sub-agents (Agent), which could be misused if malicious instructions are successfully injected via external data.
  • Sanitization: The instructions do not specify any validation or filtering mechanisms for data retrieved from web tools before it is processed by the agent.
  • Mitigations: The skill documentation explicitly instructs the agent to use environment variables for secrets and to avoid the use of real personally identifiable information (PII), reducing the potential impact of an injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 12, 2026, 08:32 PM
Security Audit — agent-trust-hub — product-launch-war-room