meta-onboarding

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly asks the agent to collect App ID/App Secret and access tokens, construct an exchange URL using them, and write those secret values verbatim into a .env file, forcing the LLM to handle and output secrets directly.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly requests Meta Marketing API access with the ads_management permission and states it can "build and manage campaigns, ads, audiences" and run Meta ads (create/update campaigns). That combination (Marketing API + ads_management + ability to build/manage campaigns) gives the agent the API capability to change ad budgets and execute spend, which qualifies as direct financial execution authority.