Answer: Handle the `tool.check` event, which is where Claude Code decides whether a tool call may run; return an object of your own such as `{ decision: 'allow' }` or `{ decision: 'deny', reason }`, or return what `await next(e)` gives you to leave the decision alone. Status: Confirmed Entry: tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call Revision: 3 Digest: sha256:438e4a682791f218 (changes whenever the content changes) Timeline: Created: 2026-11-13T03:05:14Z Last changed: 2026-11-18T12:18:54Z (next(e) decision values: — → allow) Last verified: 2026-11-20T12:23:40Z Next check: 2026-11-21T12:23:40Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of event to handle | tool.check | [1] | 2026-11-20 allow return | { decision: 'allow' } | [1] | 2026-11-20 deny return | { decision: 'deny', reason } | [1] | 2026-11-20 leave decision alone | await next(e) | [1] | 2026-11-20 next(e) decision values | allow | [1] | 2026-11-20 when tool.check fires | after the [`PreToolUse` hooks](https://code.claude.com/docs/en/hooks#pretooluse) run, and before the permission prompt | [1] | 2026-11-20 Sources: [1] Anthropic. "React to events with a mod". https://code.claude.com/docs/en/plugins/mods/events.md Role: primary. Accessed: 2026-11-20T12:23:40Z Quote: "To decide whether a tool call may run, handle [`tool.check`]" (Supports: event to handle) Quote: "return an object of your own, such as `{ decision: 'allow' }` or `{ decision: 'deny', reason }`" (Supports: allow return, deny return) Quote: "To leave the decision alone, return what `await next(e)` gives you." (Supports: leave decision alone) Quote: "Claude Code's own decision, as an object whose `decision` is `allow`, `ask`, or `deny`." (Supports: next(e) decision values) Quote: "after the [`PreToolUse` hooks](https://code.claude.com/docs/en/hooks#pretooluse) run, and before the permission prompt" (Supports: when tool.check fires) Published: 2026-11-20T03:41:00Z Snapshot SHA-256: b1a98246633f1352cfea4789628969e972a9fdb0b126114925ff98a69a589326 (captured 2026-11-20T12:23:39Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. Cross-check: not-available at 2026-11-20T12:23:40Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:23:40Z Last attempt: 2026-11-20T12:23:40Z Notes and limitations: Extracted automatically from a single official source (section "Approve or refuse a tool call before the user is asked"). Related: https://latestn.com/tool/claude-code/permissions-and-security/available-permission-modes What permission modes are available in Claude Code? https://latestn.com/tool/claude-code/permissions-and-security/available-permission-modes-2 What permission modes are available in Claude Code and what does each auto-approve? https://latestn.com/tool/claude-code/permissions-and-security/aws-credentials-sigv4-api-key How do I authenticate Claude Code with AWS credentials using SigV4 or a workspace API key? https://latestn.com/tool/claude-code/permissions-and-security/allow-mcp-tools How do I allow or auto-approve MCP tools in Claude Code Agent SDK using allowedTools? https://latestn.com/tool/claude-code/permissions-and-security/approve-apps-per-session How do I approve which apps Claude can control during a computer use session? https://latestn.com/tool/claude-code/permissions-and-security/authenticate-remote-mcp-oauth How do I authenticate Claude Code with a remote MCP server using OAuth? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-18T12:18:54Z | changed | next(e) decision values: — → allow | [1] 2026-11-15T12:23:54Z | changed | deny return: — → { decision: 'deny', reason } | [1] 2026-11-15T12:23:54Z | changed | allow return: — → { decision: 'allow' } | [1] 2026-11-15T12:23:54Z | changed | deny return value: { decision: 'deny', reason } → — | [1] 2026-11-15T12:23:54Z | changed | allow return value: { decision: 'allow' } → — | [1] 2026-11-13T03:05:14Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call POST: https://latestn.com/dispute/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). How do I approve or refuse a tool call in Claude Code before the user is asked? Revision 3, last verified 2026-11-20T12:23:40Z. https://latestn.com/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call (accessed 2026-11-21T10:03:32Z). @misc{LatestN_tool_claude_code_permissions_and_security_auto_approve_refuse_tool_call_r3, title = {How do I approve or refuse a tool call in Claude Code before the user is asked?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call}}, note = {Revision 3; digest sha256:438e4a682791f218; last verified 2026-11-20T12:23:40Z; accessed 2026-11-21T10:03:32Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/claude-code/permissions-and-security/auto-approve-refuse-tool-call