Coverage and limits

What Warden doesn’t catch

A security tool that claims to stop everything is either lying or a sandbox. Warden is neither. Here is where its view ends, so you can decide what else you need.

Coverage per agent

What each agent’s hook system lets Warden check before the action runs.

Which actions can be checked before they run, per agent
AgentShell commandsFile readsFile writesMCP toolsNative “ask”
Claude CodeYesYesYesYesYes
Codex CLIYesThrough shell parsingYes (apply_patch)YesNo, Warden’s popup
CursorYesYesPartly (blocked at tool level; edit details arrive after)YesShell and MCP
Gemini CLIYesYesYesYesNo, Warden’s popup
WindsurfYesYesYesYesNo, Warden’s popup
ClineYesYesYesYesNo, Warden’s popup

Cline checks only run after you switch the hook on in Cline’s Hooks tab. Not possible with hooks: Aider and other agents that have no pre-action hook.

Limits every hook-based guard shares

When something fails

Agents treat a crashing hook differently. Claude Code and Codex let the action through if a hook errors, so Warden’s helper is written to always answer with an explicit decision. Cursor hooks are installed with failClosed, so a failure blocks. If the menu-bar app isn’t running, hard blocks still apply; see the FAQ for how questions are handled.

Found a gap that isn’t listed? If it’s a working way around a rule, report it privately on GitHub; bypass reports are the most useful thing you can send us. For anything else, open an issue or email ali@dakcyber.com. We’ll add it here, fix it, or both.