conference-networking-prep

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly instructs the agent to call enrichment endpoints with reveal_email (and optionally reveal_phone_numbers) enabled and to include those contact details in lead lists and per-person cards, which requires the LLM to handle and output sensitive contact data verbatim.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The skill uses WebFetch to extract free-form text from outsider-authored public conference pages (e.g., speakers/sponsors/agenda/session descriptions) and then includes that extracted prose in the agent’s LLM context to generate the playbook, creating an indirect prompt-injection exposure path.