https://thomasdevos.com/tags/promptinjection/ Recent content in Promptinjection on Thomas Talks AI: Claude Architecture, AI Agent Security, Claude Code & Production Engineering Hugo -- 0.145.0 en-us Thomas De Vos © Mon, 17 Aug 2026 10:43:58 +0100 https://thomasdevos.com/posts/2026-08-17.10.43.58-do-not-let-mcp-tool-output-rewrite-claude-codes-instructions/ Mon, 17 Aug 2026 10:43:58 +0100 https://thomasdevos.com/posts/2026-08-17.10.43.58-do-not-let-mcp-tool-output-rewrite-claude-codes-instructions/ Learn how to defend Claude Code against indirect prompt injection in MCP tool output with typed responses, taint labels, capability checks, denial evals, and review traces.