agentmail

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill's examples and SDK usage explicitly show embedding an API key as "Authorization: Bearer am_..." and as apiKey: "am_...", which instructs including secret values verbatim in curl headers and code (an LLM following the skill would need to insert real secrets into outputs/requests).

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md describes runtime reading of inbound email via the API (e.g., “Read inbox” and “Process reply” using detail.bodyText/detail.bodyHtml), which would ingest outsider-authored message bodies into the agent’s LLM context if the agent relays/uses those fields for prompting.