---
title: "Terminal commands"
canonical: "https://canmyagentuse.com/features/terminal"
contentKind: "feature"
locale: "en"
description: "Propose or run shell commands in a project environment."
llmSummary: "Terminal command support lets a product propose or execute shell commands in a project environment. Approval, shell, sandbox, working directory, and persistence are qualifiers."
publishedAt: "2026-08-28T00:00:00.000Z"
updatedAt: "2026-08-28T00:00:00.000Z"
verifiedAt: "2026-08-28"
tags: ["workspace","terminal"]
---

# Terminal commands

Terminal command support lets a product propose or execute shell commands in a project environment. Approval, shell, sandbox, working directory, and persistence are qualifiers.

- HTML: https://canmyagentuse.com/features/terminal
- JSON: https://canmyagentuse.com/api/v1/features/terminal.json
- Markdown: https://canmyagentuse.com/features/terminal.md

Terminology basis: **Common product term** — https://docs.anthropic.com/en/docs/claude-code/cli-usage.

## Current support at a glance

Terminal commands: 14 supported, 1 partial, 0 unsupported, 16 unreviewed across 31 cataloged products.

- Reviewed current products: 15 of 31
- Supported: 14
- Partial: 1
- Unsupported: 0
- Unreviewed: 16
- Not applicable: 0

Unknown or unreviewed means insufficient published evidence; it does not mean unsupported.

This capability covers a built-in, permissioned route for an agent to propose or execute shell commands in its working environment. It does not claim unattended execution, access beyond the selected workspace, a particular shell, or identical approval behavior.

Only current product documentation is used here. Preview tracks and unreviewed surfaces remain unknown.

## Catalog context

- Category: [tools](/categories/tools.md)
- Terminology basis: Common product term
- Aliases: None

## Compatibility assertions

Unknown means insufficient published evidence; it does not mean unsupported.

### ChatGPT (web)

- Harness: [ChatGPT](/harnesses/chatgpt-web.md)
- current: **Unknown**
- preview: **Unknown**

### Claude (web)

- Harness: [Claude](/harnesses/claude-web.md)
- current: **Unknown**
- preview: **Unknown**

### Gemini (web)

- Harness: [Gemini](/harnesses/gemini-web.md)
- current: **Unknown**
- preview: **Unknown**

### Copilot (web)

- Harness: [Copilot](/harnesses/copilot-web.md)
- current: **Unknown**
- preview: **Unknown**

### Grok (web)

- Harness: [Grok](/harnesses/grok-web.md)
- current: **Unknown**
- preview: **Unknown**

### Grok Bot (desktop)

- Harness: [Grok Bot](/harnesses/grok-bot-desktop.md)
- current: **Supported**
  - Target: hosted-observation — 2026-08-28 Grok Bot desktop documentation observation; observed 2026-08-28
  - Environment: hosted-default
  - Constraint (runtime): commands run on the account-scoped hosted computer by default; credentials and files there are shared across the account's Bots
  - Constraint (policy): commands on the operator's local Mac or Windows computer are separately controlled and default to asking every time
  - Evidence: [xAI — Grok Bot computer and apps](https://docs.x.ai/grok-bot/computer-and-apps) — documented; observed 2026-08-28
  - Qualification note 7: Evidence checked 2026-08-28: Grok Bot's persistent cloud computer includes a command line and filesystem; local-computer commands are a separate capability controlled by a desktop approval policy.

### Perplexity (web)

- Harness: [Perplexity](/harnesses/perplexity-web.md)
- current: **Unknown**
- preview: **Unknown**

### Le Chat (web)

- Harness: [Le Chat](/harnesses/le-chat.md)
- current: **Unknown**
- preview: **Unknown**

### Devin (web)

- Harness: [Devin](/harnesses/devin-web.md)
- current: **Supported**
  - Target: hosted-observation — 2026-08-28 Devin documentation observation; observed 2026-08-28
  - Environment: hosted-default
  - Constraint (runtime): commands run in Devin's hosted development environment rather than the operator's local terminal
  - Evidence: [Devin — Session tools](https://docs.devin.ai/work-with-devin/devin-session-tools) — documented; observed 2026-08-28
  - Qualification note 14: Evidence checked 2026-08-28: Devin executes shell commands in its hosted development environment and exposes the complete command history and outputs in the session UI.
- preview: **Unknown**

### Replit Agent (web)

- Harness: [Replit Agent](/harnesses/replit-agent.md)
- current: **Unknown**
- preview: **Unknown**

### ChatGPT (desktop)

- Harness: [ChatGPT](/harnesses/chatgpt-desktop.md)
- current: **Unknown**
- preview: **Unknown**

### Claude (desktop)

- Harness: [Claude](/harnesses/claude-desktop.md)
- current: **Unknown**
- preview: **Unknown**

### Cursor (desktop)

- Harness: [Cursor](/harnesses/cursor.md)
- current: **Supported**
  - Target: dated-documentation — current Cursor documentation; observed 2026-08-28
  - Environment: local-default
  - Evidence: [Cursor — Agent overview](https://cursor.com/docs/agent/overview) — documented; observed 2026-08-28
  - Qualification note 2: Evidence checked 2026-08-28: Cursor Agent documents running terminal commands while completing coding tasks.
- preview: **Unknown**

### OpenWork Desktop (desktop)

- Harness: [OpenWork Desktop](/harnesses/openwork-desktop.md)
- current: **Supported**
  - Target: dated-documentation — current OpenWork architecture documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): shell execution remains subject to OpenWork and OpenCode permission rules
  - Evidence: [OpenWork — Architecture](https://github.com/different-ai/openwork/blob/dev/ARCHITECTURE.md) — documented; observed 2026-08-28
  - Qualification note 6: Evidence checked 2026-08-28: OpenWork's OpenCode-backed runtime includes shell execution, with permission requests surfaced in the desktop app and policy controlled by the host.

### Copilot Chat (desktop)

- Harness: [Copilot Chat](/harnesses/vscode-copilot.md)
- current: **Supported**
  - Target: dated-documentation — current VS Code documentation; observed 2026-08-28
  - Environment: local-default
  - Evidence: [Visual Studio Code — Tools in agents](https://code.visualstudio.com/docs/copilot/concepts/tools) — documented; observed 2026-08-28
  - Qualification note 5: Evidence checked 2026-08-28: VS Code documents terminal tools for agent mode, with user-configurable approval and auto-approval behavior.
- preview: **Unknown**

### Chrome WebMCP origin trial (desktop)

- Harness: [Chrome WebMCP origin trial](/harnesses/chrome-webmcp-preview.md)
- current: **Unknown**

### Windsurf (desktop)

- Harness: [Windsurf](/harnesses/windsurf.md)
- current: **Unknown**
- preview: **Unknown**

### Zed Agent (desktop)

- Harness: [Zed Agent](/harnesses/zed-agent.md)
- current: **Supported**
  - Target: dated-documentation — current Zed documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): terminal calls remain subject to Zed Agent profiles, tool permissions, and workspace trust
  - Evidence: [Zed — AI quick start](https://zed.dev/docs/ai/quick-start) — documented; observed 2026-08-28
  - Qualification note 12: Evidence checked 2026-08-28: Zed's native Agent is documented as able to run code in the project, and its permission system includes a terminal tool.
- preview: **Unknown**

### Continue (desktop)

- Harness: [Continue](/harnesses/continue.md)
- current: **Supported**
  - Target: dated-documentation — current Continue documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): terminal execution is available in Agent mode, not read-only Plan mode, and depends on model tool-use support and tool policy
  - Evidence: [Continue — How Agent mode works](https://docs.continue.dev/features/agent/how-it-works) — documented; observed 2026-08-28
  - Qualification note 10: Evidence checked 2026-08-28: Continue Agent mode exposes `run_terminal_command`, which runs commands from the workspace root after the tool-permission step unless configured Automatic.
- preview: **Unknown**

### Cline (desktop)

- Harness: [Cline](/harnesses/cline.md)
- current: **Supported**
  - Target: dated-documentation — current Cline documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): approval, auto-approval, or disablement depends on Cline settings and policy
  - Evidence: [Cline — Tools reference](https://docs.cline.bot/tools-reference/all-cline-tools) — documented; observed 2026-08-28
  - Qualification note 9: Evidence checked 2026-08-28: Cline's built-in Bash tool runs commands in a persistent shell session and can be approval-gated by settings or policy.
- preview: **Unknown**

### JetBrains AI (desktop)

- Harness: [JetBrains AI](/harnesses/jetbrains-ai.md)
- current: **Unknown**
- preview: **Unknown**

### Warp (desktop)

- Harness: [Warp](/harnesses/warp.md)
- current: **Supported**
  - Target: dated-documentation — current Warp documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): the local Agent documents review and approval controls before actions execute
  - Evidence: [Warp — Getting started with Warp](https://docs.warp.dev/) — documented; observed 2026-08-28
  - Qualification note 13: Evidence checked 2026-08-28: Warp Agent runs commands and interprets results in the local app; the same agent is also available through Warp Agent CLI.
- preview: **Unknown**

### Claude CLI (cli)

- Harness: [Claude CLI](/harnesses/claude-cli.md)
- current: **Supported**
  - Target: dated-documentation — current Claude Code CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Evidence: [Anthropic — Claude Code CLI reference](https://docs.anthropic.com/en/docs/claude-code/cli-usage) — documented; observed 2026-08-28
  - Qualification note 1: Evidence checked 2026-08-28: Claude Code documents Bash as a permission-controlled tool and provides CLI modes for agentic command execution.
- preview: **Unknown**

### ChatGPT CLI (cli)

- Harness: [ChatGPT CLI](/harnesses/chatgpt-cli.md)
- current: **Unknown**
- preview: **Unknown**

### Codex CLI (cli)

- Harness: [Codex CLI](/harnesses/codex-cli.md)
- current: **Supported**
  - Target: dated-documentation — current Codex CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Evidence: [OpenAI — Get started with Codex CLI](https://help.openai.com/en/articles/11096431) — documented; observed 2026-08-28
  - Qualification note 3: Evidence checked 2026-08-28: OpenAI documents Codex CLI as able to run code and commands locally in the selected directory.
- preview: **Unknown**

### OpenCode (cli)

- Harness: [OpenCode](/harnesses/opencode.md)
- current: **Supported**
  - Target: dated-documentation — current OpenCode documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): the Bash tool can be allowed, approval-gated, or denied through permission rules
  - Evidence: [OpenCode — Tools](https://opencode.ai/docs/tools/) — documented; observed 2026-08-28
  - Qualification note 11: Evidence checked 2026-08-28: OpenCode's built-in Bash tool lets the model execute shell commands in the project environment, subject to configurable allow, ask, or deny permissions.
- preview: **Unknown**

### Gemini CLI (cli)

- Harness: [Gemini CLI](/harnesses/gemini-cli.md)
- current: **Supported**
  - Target: dated-documentation — current Gemini CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Evidence: [Gemini CLI — Tool reference](https://github.com/google-gemini/gemini-cli/blob/main/docs/reference/tools.md) — documented; observed 2026-08-28
  - Qualification note 4: Evidence checked 2026-08-28: Gemini CLI's official tool reference includes the `run_shell_command` tool with workspace and approval controls.
- preview: **Unknown**

### Aider (cli)

- Harness: [Aider](/harnesses/aider.md)
- current: **Partial**
  - Target: dated-documentation — current Aider documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): the documented /run command is invoked by the person; this evidence does not establish an autonomous agent shell tool
  - Evidence: [Aider — In-chat commands](https://aider.chat/docs/usage/commands.html) — documented; observed 2026-08-28
  - Qualification note 15: Evidence checked 2026-08-28: Aider includes a user-invoked `/run` command for executing shell commands and optionally adding output to chat, but the cited reference does not establish an autonomous model tool for proposing and running commands.
- preview: **Unknown**

### Goose (cli)

- Harness: [Goose](/harnesses/goose.md)
- current: **Unknown**
- preview: **Unknown**

### Copilot CLI (cli)

- Harness: [Copilot CLI](/harnesses/copilot-cli.md)
- current: **Supported**
  - Target: dated-documentation — current GitHub Copilot CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (policy): shell execution is scoped by repository or working-directory permissions and potentially destructive commands require approval unless pre-authorized
  - Evidence: [GitHub — Allowing and denying tool use in Copilot CLI](https://docs.github.com/en/copilot/how-tos/copilot-cli/use-copilot-cli/allowing-tools) — documented; observed 2026-08-28
  - Qualification note 8: Evidence checked 2026-08-28: GitHub Copilot CLI can run shell tools in the current repository or directory, with potentially destructive commands gated by its permission system.
- preview: **Unknown**

### Amp (cli)

- Harness: [Amp](/harnesses/amp-cli.md)
- current: **Unknown**
- preview: **Unknown**