Usage: copilot [OPTIONS] [COMMAND]

GitHub Copilot CLI - An AI-powered coding assistant.

Start an interactive session to chat with Copilot, or use -p/--prompt for
non-interactive scripting. Copilot can edit files, run shell commands, search
your codebase, and more — all with configurable permissions.

Run `copilot <command> --help` for details on any subcommand.

Commands:
  app          Open the GitHub Copilot app
  login        Authenticate with Copilot
  help         Display help information
  init         Initialize Copilot instructions
  config       Manage configuration settings
  update       Download the latest version
  version      Display version information
  workflow     Run dynamic workflows
  sessions     Manage saved sessions
  memories     Manage memories
  plugin       Manage plugins
  mcp          Manage MCP servers
  skill        Manage skills
  instruction  Inspect instruction sources
  lsp          Inspect language server configuration
  sandbox      Manage command sandboxing
  completion   Generate a shell completion script

Options:
  -v, --version
          show version information
  -i, --interactive <prompt>
          Start interactive mode and automatically execute this prompt
      --fleet
          Run the prompt in fleet mode (parallel subagent
          orchestration); combine with -i, -p, or piped stdin
  -p, --prompt <text>
          Execute a prompt in non-interactive mode (exits after
          completion)
  -s, --silent
          Output only the agent response (no stats), useful for
          scripting with -p
      --enable-memory
          Enable memory in prompt mode (disabled by default)
      --model <model>
          Set the AI model to use (use 'auto' to let Copilot pick
          automatically)
      --reasoning-effort <level>
          Set the reasoning effort level [possible values: none, minimal, low, medium, high, xhigh, max]
      --context <tier>
          Set the context window tier (overrides persisted setting) [possible values: default, long_context]
      --auto-tier <preference>
          Set the Auto routing profile (validated against the available catalog)
      --agent <agent>
          Specify a custom agent to use
  -r, --resume [<value>]
          Resume from a previous session (optionally specify existing
          session ID, task ID, ID prefix, or name; name matching is
          exact, case-insensitive)
      --continue
          Resume the most recent session
  -n, --name <name>
          Set a name for the new session
      --session-id <id>
          Resume an existing session or task by ID, or set the UUID
          for a new session
      --connect [<sessionId>]
          Connect directly to a remote session (optionally specify
          session ID or task ID)
      --allow-all-tools
          Allow all tools to run automatically without confirmation;
          required for non-interactive mode. Run `copilot help environment`
          for what COPILOT_ALLOW_ALL accepts and what exact "true" grants [env: COPILOT_ALLOW_ALL=]
      --allow-all-paths
          Disable file path verification and allow access to any path
      --disallow-temp-dir
          Prevent automatic access to the system temporary directory
      --banner
          Show the startup banner
      --screen-reader
          Enable screen reader optimizations
      --plain-diff
          Disable rich diff rendering (syntax highlighting via diff
          tool specified by git config)
  -C <directory>
          Change working directory before doing anything else
      --log-dir <directory>
          Set log file directory (default: ~/.copilot/logs/)
      --extension-sdk-path <directory>
          Override the bundled @github/copilot-sdk injected into
          extension subprocesses with a local `copilot-sdk/` folder.
          Invalid paths fall back to the bundled SDK.
      --log-level <level>
          Set the log level [possible values: none, error, warning, info, debug, all, default]
      --stream <mode>
          Enable or disable streaming mode [possible values: on, off]
      --output-format <format>
          Output format: 'text' (default) or 'json' (JSONL, one JSON
          object per line) [possible values: text, json]
      --share [<path>]
          Share session to markdown file after completion in
          non-interactive mode (default: ./copilot-session-<id>.md)
      --share-gist
          Share session to a secret GitHub gist after completion in
          non-interactive mode
      --add-dir <directory>
          Allow file access to a directory and load its .github/skills
          and .github/agents as trusted configuration (can be used
          multiple times). A relative path resolves against the
          session working directory (the --resume/--worktree/-C
          directory), whatever order the options appear in
      --attachment <path>
          Attach a file (image or native document) to the initial
          prompt; only valid in non-interactive mode (can be used
          multiple times)
      --disable-mcp-server <server-name>
          Disable a specific MCP server (can be used multiple times)
      --disable-builtin-mcps
          Disable all built-in MCP servers (currently:
          github-mcp-server, githubiq)
      --enable-all-github-mcp-tools
          Enable all GitHub MCP server tools instead of the default
          CLI subset. Overrides --add-github-mcp-toolset and
          --add-github-mcp-tool options.
      --add-github-mcp-toolset <toolset>
          Add a toolset to enable for the GitHub MCP server instead of
          the default CLI subset (can be used multiple times). Use
          "all" for all toolsets.
      --add-github-mcp-tool <tool>
          Add a tool to enable for the GitHub MCP server instead of
          the default CLI subset (can be used multiple times). Use "*"
          for all tools.
      --plugin-dir <directory>
          Load a plugin from a local directory (can be used multiple
          times). A relative path resolves against the session working
          directory (the --resume/--worktree/-C directory), whatever
          order the options appear in
      --additional-mcp-config <json>
          Additional MCP servers configuration as JSON string or file
          path (prefix with @) (can be used multiple times; augments
          config from ~/.copilot/mcp-config.json for this session). A
          relative @ path resolves against the session working
          directory (the --resume/--worktree/-C directory) and a
          leading ~/ expands to your home directory
      --mcp-github-auth <server=origin>
          Send the signed-in GitHub account credential only to this
          explicit --additional-mcp-config server and approved origin.
          HTTPS is required except for literal loopback HTTP. Never
          put a token in this argument. Can be used multiple times.
      --allow-all-mcp-server-instructions
          Include initialization instructions from all MCP servers in
          the system prompt instead of only allowlisted servers
      --allow-tool [<tools>...]
          Tools the CLI has permission to use; will not prompt for
          permission
      --deny-tool [<tools>...]
          Tools the CLI does not have permission to use; will not
          prompt for permission
      --available-tools [<tools>...]
          Only these tools will be available to the model
      --excluded-tools [<tools>...]
          These tools will not be available to the model
      --secret-env-vars [<vars>...]
          Environment variable names whose values are stripped from
          shell and MCP server environments and redacted from output
          (e.g., --secret-env-vars=MY_KEY,OTHER_KEY)
      --allow-url [<urls>...]
          Allow access to specific URLs or domains
      --deny-url [<urls>...]
          Deny access to specific URLs or domains, takes precedence
          over --allow-url
      --allow-all-urls
          Allow access to all URLs without confirmation
      --allow-all
          Enable all permissions (equivalent to --allow-all-tools
          --allow-all-paths --allow-all-urls)
      --yolo
          Enable all permissions (equivalent to --allow-all-tools
          --allow-all-paths --allow-all-urls)
      --max-autopilot-continues <count>
          Maximum number of continuation messages in autopilot mode [default: 5]
      --mode <mode>
          Set the initial agent mode [possible values: interactive, plan, autopilot]
      --autopilot
          Start in autopilot mode
      --plan
          Start in plan mode (combine with --mode autopilot to
          auto-approve the plan and implement it autonomously)
      --experimental
          Enable experimental features
      --bash-env [<value>]
          Enable BASH_ENV support for bash shells (on|off)
      --mouse [<value>]
          Enable mouse support in alt screen mode (on|off)
      --acp
          Start as Agent Client Protocol server
      --remote
          Enable remote control of your session from GitHub web and
          mobile
      --remote-export
          Export your session to GitHub web and mobile (read-only;
          does not enable remote control)
      --dynamic-retrieval <category=on|off>
          Enable or disable embeddings-based dynamic retrieval per
          category and persist the choice (category: skills).
          Repeatable, e.g. --dynamic-retrieval skills=off
      --enable-mcp-server <server-name>
          Enable an MCP server disabled in settings for this run only;
          nothing is persisted (can be used multiple times)
      --max-ai-credits <credits>
          Set max AI credits for this session
      --assisted-approval
          Review tool permission requests with the assisted-approval
          safety judge instead of approving them outright. The judge
          is probabilistic and not a security boundary. Equivalent
          to the `assisted` mode of /permissions. Takes precedence
          over --allow-all-tools when the judge engages; requires
          --experimental or enabledFeatureFlags.AUTO_APPROVAL. (env:
          COPILOT_ASSISTED_APPROVAL)
      --usage-output-file <file>
          Write final usage statistics as JSON to the specified file
      --no-custom-instructions
          Disable loading of custom instructions from AGENTS.md and
          related files
      --no-auto-update
          Disable downloading CLI updates automatically. Copilot then
          runs the version bundled in the executable you launched
          rather than a newer previously downloaded one
      --no-ask-user
          Disable the ask_user tool (agent works autonomously without
          asking questions)
      --no-color
          Disable all color output
      --no-experimental
          Disable experimental features
      --no-bash-env
          Disable BASH_ENV support for bash shells
      --no-mouse
          Disable mouse support in alt screen mode
      --no-remote
          Disable remote control of your session from GitHub web and
          mobile
      --no-remote-export
          Disable exporting your session to GitHub web and mobile
          (also disables remote control)
      --no-eager-powershell-resolution
          Disable background PowerShell prompt resolution on Windows
  -h, --help
          Print help

Help Topics:
  billing      AI credit usage
  commands     Interactive Mode Commands
  config       Configuration Settings
  environment  Environment Variables
  limits       Session Limits Controls
  logging      Logging
  monitoring   Monitoring with OpenTelemetry
  permissions  Permissions
  providers    Custom Model Providers (BYOK)
  sandbox      Command Sandboxing

Examples:
  # Start interactive mode
  $ copilot

  # Start interactive mode and automatically execute a prompt
  $ copilot -i "Fix the bug in main.js"

  # Execute a prompt in non-interactive mode (exits after completion)
  $ copilot -p "Fix the bug in main.js" --allow-all-tools

  # Enable all permissions with a single flag
  $ copilot -p "Fix the bug in main.js" --allow-all
  $ copilot -p "Fix the bug in main.js" --yolo

  # Start with a specific model
  $ copilot --model gpt-5.4

  # Resume the most recent session
  $ copilot --continue

  # Resume a previous session using session picker
  $ copilot --resume

  # Resume a specific session by ID
  $ copilot --resume=<session-id>

  # Name a new session
  $ copilot --name="my feature"

  # Start a new session with a specific UUID
  $ copilot --session-id=0cb916db-26aa-40f2-86b5-1ba81b225fd2

  # Resume or attach to an existing session or task by ID
  $ copilot --session-id=<session-or-task-id>

  # Resume a session by name
  $ copilot --resume="my feature"

  # Resume a session by ID prefix (7+ hex chars)
  $ copilot --resume=0cb916d

  # Resume with auto-approval
  $ copilot --allow-all-tools --resume

  # Allow access to additional directory
  $ copilot --add-dir /home/user/projects

  # Allow multiple directories
  $ copilot --add-dir ~/workspace --add-dir /tmp

  # Disable path verification (allow access to any path)
  $ copilot --allow-all-paths

  # Allow all git commands except git push
  $ copilot --allow-tool='shell(git:*)' --deny-tool='shell(git push)'

  # Allow all file editing
  $ copilot --allow-tool='write'

  # Allow all but one specific tool from MCP server with name "MyMCP"
  $ copilot --deny-tool='MyMCP(denied_tool)' --allow-tool='MyMCP'

  # Allow GitHub API access (defaults to HTTPS)
  $ copilot --allow-url=github.com

  # Deny access to specific domain over HTTPS
  $ copilot --deny-url=https://malicious-site.com
  $ copilot --deny-url=malicious-site.com

  # Allow all URLs without confirmation
  $ copilot --allow-all-urls

  # Initialize Copilot instructions for a repository
  $ copilot init

Learn More:
  Use `copilot <command> --help` for more information about a command.
  Read the documentation at https://docs.github.com/copilot/how-tos/copilot-cli