Elicitation Hook - Claude Code
Fires when an MCP server requests input from the user.
The Elicitation hook intercepts when an MCP server asks the user a question. The hook can provide the answer automatically or pass through to a prompt.
What it does
MCP servers sometimes need additional input - a confirmation, a choice, a free-text value. Normally Claude surfaces the request as a dialog. The Elicitation hook can answer it programmatically based on the question, the server, or the current state, skipping the dialog and keeping the session flowing.
When to use it
- Auto-answering routine MCP confirmations in a trusted context.
- Keeping headless sessions from hanging on MCP prompts.
- Team policies where a specific MCP server's questions have a canonical answer.
- Integrating MCP workflows with your own approval backend.
Gotchas
- Auto-answering without care defeats the MCP server's safety intent. Scope carefully.
- Not every elicitation has a sensible automated response. Some must reach a human.
- Logging answers is important for audit - you're effectively speaking for the user.
Official docs: https://code.claude.com/docs/en/hooks.md#elicitation
Technical content at the intersection of AI and development. Building with AI agents, Claude Code, and modern dev tools - then showing you exactly how it works.
Get the weekly deep dive
Tutorials on Claude Code, AI agents, and dev tools - delivered free every week.
Was this helpful?




