linear

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

• Third-party content exposure detected (high risk: 0.90). The skill programmatically fetches and parses user-authored Linear issue descriptions, comments, and attachments via the Linear GraphQL/MCP flows (see SKILL.md and references/epic-format.md), and explicitly uses user-provided ## Instructions (exposed to agents as EPIC_INSTRUCTIONS) and public GitHub marketplace.json for validation—all of which are untrusted, user-generated content that the agent is instructed to read and act on.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill issues runtime requests to https://api.linear.app/graphql (and can use the MCP endpoint https://mcp.linear.app/mcp) to fetch epic descriptions including ## Instructions which are injected as EPIC_INSTRUCTIONS and directly influence agent prompts, so these endpoints are runtime remote-content dependencies that can change agent behavior.