Answer: Gemini CLI includes confirmation prompts requiring explicit approval for every shell command by default, and optional sandboxing that runs all shell commands inside a secure Docker container. Sandboxing is enabled with the `--sandbox` flag. Status: Confirmed Entry: tool/gemini-cli/permissions-and-security/shell-command-safety-features Revision: 1 Digest: sha256:2887a142b3e5f9f3 (changes whenever the content changes) Timeline: Created: 2026-11-13T03:18:03Z Last changed: 2026-11-13T03:18:03Z Last verified: 2026-11-20T12:52:04Z Next check: 2026-11-21T12:52:04Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of Confirmation prompts | every | [1] | 2026-11-20 Approval options | Allow once | [1] | 2026-11-20 Sandboxing | Docker container | [1] | 2026-11-20 Sandbox flag | --sandbox | [1] | 2026-11-20 Sources: [1] Google. "Execute shell commands". https://geminicli.com/docs/cli/tutorials/shell-commands.md Role: primary. Accessed: 2026-11-20T12:52:04Z Quote: "By default, **every** shell command requested by the agent requires your" (Supports: Confirmation prompts) Quote: "**Allow once:** Runs the command one time." (Supports: Approval options) Quote: "This runs all shell commands inside a secure Docker container." (Supports: Sandboxing) Quote: "Use the `--sandbox` flag when starting the CLI:" (Supports: Sandbox flag) Published: 2026-10-16T02:36:07Z Snapshot SHA-256: 8cca18dbe8136e49b1038353ce8423a9e1def567d7aec364e8e42da95a5f569a (captured 2026-11-13T03:18:02Z) 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 2026-11-13T03:18:03Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T12:52:04Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:52:04Z Last attempt: 2026-11-20T12:52:04Z Notes and limitations: Extracted automatically from a single official source (section "Safety features"). Related: https://latestn.com/tool/gemini-cli/permissions-and-security/skill-activation-consent How does Gemini CLI activate an Agent Skill and ask for consent? https://latestn.com/tool/gemini-cli/permissions-and-security/skill-security-privacy What security and privacy considerations should I follow when creating Gemini CLI Agent Skills? https://latestn.com/tool/gemini-cli/permissions-and-security/tool-execution-security How does Gemini CLI handle confirmation and sandboxing for tool execution? https://latestn.com/tool/gemini-cli/permissions-and-security/sandbox-proxy-script-setup How do I configure a proxy script for the Gemini CLI sandbox using GEMINI_SANDBOX_PROXY_COMMAND? https://latestn.com/tool/gemini-cli/permissions-and-security/sandboxing-methods Which sandboxing methods does Gemini CLI support on macOS, Linux, and Windows? https://latestn.com/tool/gemini-cli/permissions-and-security/secure-api-key-storage What is the best way to store my API keys securely for Gemini CLI? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-13T03:18:03Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/gemini-cli/permissions-and-security/shell-command-safety-features/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/gemini-cli/permissions-and-security/shell-command-safety-features POST: https://latestn.com/dispute/tool/gemini-cli/permissions-and-security/shell-command-safety-features with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/gemini-cli/permissions-and-security/shell-command-safety-features?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What safety features like confirmation prompts and sandboxing does Gemini CLI have for shell commands? Revision 1, last verified 2026-11-20T12:52:04Z. https://latestn.com/tool/gemini-cli/permissions-and-security/shell-command-safety-features (accessed 2026-11-21T10:00:06Z). @misc{LatestN_tool_gemini_cli_permissions_and_security_shell_command_safety_features_r1, title = {What safety features like confirmation prompts and sandboxing does Gemini CLI have for shell commands?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/gemini-cli/permissions-and-security/shell-command-safety-features}}, note = {Revision 1; digest sha256:2887a142b3e5f9f3; last verified 2026-11-20T12:52:04Z; accessed 2026-11-21T10:00:06Z} } 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/gemini-cli/permissions-and-security/shell-command-safety-features