lsp-fix-all

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

• Third-party content exposure detected (medium risk: 0.65). The skill ingests outsider-authored free text via mcp__lsp__get_diagnostics and mcp__lsp__suggest_fixes, which return LSP-provided diagnostic messages and code-action titles/content derived from the target file’s contents (i.e., text the operating user did not author), and those messages are then placed into the agent’s LLM context for filtering/looping.