Answer: By default, sensitive actions require manual approval: configuration files need your approval before changes, terminal commands need your approval, and all MCP connections and each MCP tool call need approval. Reading files and searching code don't require approval. Status: Confirmed Entry: tool/cursor/permissions-and-security/default-approval-actions Revision: 1 Digest: sha256:41cf281cc28918fc (changes whenever the content changes) Timeline: Created: 2026-11-13T03:09:17Z Last changed: 2026-11-13T03:09:17Z Last verified: 2026-11-20T12:34:48Z Next check: 2026-11-21T12:34:48Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of sensitive actions | sensitive actions require your manual approval | [1] | 2026-11-20 configuration files | Configuration files (like workspace settings) need your approval first. | [1] | 2026-11-20 terminal commands | terminal commands need your approval | [1] | 2026-11-20 MCP connections | All MCP connections need your approval. | [1] | 2026-11-20 MCP tool calls | each tool call still needs individual approval before running | [1] | 2026-11-20 no approval needed | Reading files and searching code don't require approval. | [1] | 2026-11-20 Sources: [1] Anysphere. "cursor.com/docs/agent/security.md". https://cursor.com/docs/agent/security.md Role: primary. Accessed: 2026-11-20T12:34:48Z Quote: "By default, sensitive actions require your manual approval." (Supports: sensitive actions) Quote: "Configuration files (like workspace settings) need your approval first." (Supports: configuration files) Quote: "By default, terminal commands need your approval." (Supports: terminal commands) Quote: "All MCP connections need your approval." (Supports: MCP connections) Quote: "After you approve an MCP connection, each tool call still needs individual approval before running." (Supports: MCP tool calls) Quote: "Reading files and searching code don't require approval." (Supports: no approval needed) Snapshot SHA-256: d54ccd9897c20d5c08b235de8f3adbf1bcc30a95d072988de6a2b773abff9572 (captured 2026-11-13T03:09:17Z) 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]. On this check the text of [1] was unchanged since 2026-11-13T03:09:17Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T12:34:48Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:34:48Z Last attempt: 2026-11-20T12:34:48Z Notes and limitations: Extracted automatically from a single official source (section "Agent Security"). Related: https://latestn.com/tool/cursor/permissions-and-security/enable-oidc-self-hosted-workers How do I enable OIDC token minting for Cursor Cloud Agents on self-hosted workers? https://latestn.com/tool/cursor/permissions-and-security/enforcement-hooks How do I set up enforcement hooks in Cursor to block or redact operations? https://latestn.com/tool/cursor/permissions-and-security/file-read-permissions How do I restrict file read access in Cursor CLI with Read permission patterns? https://latestn.com/tool/cursor/permissions-and-security/configure-cli-permissions How do I configure permissions for the Cursor Agent CLI? https://latestn.com/tool/cursor/permissions-and-security/configure-origin-allowlist How do I configure the origin allowlist for Cursor's browser? https://latestn.com/tool/cursor/permissions-and-security/configure-sandbox How do I configure sandboxing and read access for the Cursor agent? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-13T03:09:17Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/cursor/permissions-and-security/default-approval-actions/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/cursor/permissions-and-security/default-approval-actions POST: https://latestn.com/dispute/tool/cursor/permissions-and-security/default-approval-actions with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/cursor/permissions-and-security/default-approval-actions?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). Which Cursor agent actions require my manual approval by default? Revision 1, last verified 2026-11-20T12:34:48Z. https://latestn.com/tool/cursor/permissions-and-security/default-approval-actions (accessed 2026-11-21T10:01:55Z). @misc{LatestN_tool_cursor_permissions_and_security_default_approval_actions_r1, title = {Which Cursor agent actions require my manual approval by default?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/cursor/permissions-and-security/default-approval-actions}}, note = {Revision 1; digest sha256:41cf281cc28918fc; last verified 2026-11-20T12:34:48Z; accessed 2026-11-21T10:01:55Z} } 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/cursor/permissions-and-security/default-approval-actions