sales-do

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

• Third-party content exposure detected (high risk: 1.00). The router explicitly instructs (Step 2.5 "How to find the skill file") to WebFetch SKILL.md from raw.githubusercontent.com (public GitHub), and those user-authored SKILL.md files are read and used to decide routing and generate prompts, exposing the agent to untrusted third‑party content that can influence actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the router to WebFetch SKILL.md files at runtime from raw.githubusercontent.com (e.g. https://raw.githubusercontent.com/sales-skills/sales/main/skills/{skill}/SKILL.md), and those fetched files are read and used to build the agent's prompts, so this is a runtime external dependency that directly controls agent instructions.