Skip to main content
The Deep Agents CLI is an open source terminal coding agent built on the Deep Agents SDK. It retains persistent memory, maintains context across sessions, learns project conventions, uses customizable skills, and executes code with approval controls. Deep Agents CLI The Deep Agents CLI has the following built-in capabilities:
  • File operations - read, write, and edit files in your project with tools that enable agents to manage and modify code and documentation.
  • Shell command execution - execute shell commands to run tests, build projects, manage dependencies, and interact with version control systems.
  • Web search - search the web for up-to-date information and documentation (requires Tavily API key).
  • HTTP requests - make HTTP requests to APIs and external services for data fetching and integration tasks.
  • Task planning and tracking - break down complex tasks into discrete steps and track progress through the built-in todo system.
  • Memory storage and retrieval - store and retrieve information across sessions, enabling agents to remember project conventions and learned patterns.
  • Human-in-the-loop - require human approval for sensitive tool operations.
  • Skills - extend agent capabilities with custom expertise and instructions stored in skill directories.
  • MCP tools - load external tools from Model Context Protocol servers via auto-discovered or explicit config files.

​
Built-in tools

The agent comes with the following built-in tools which are available without configuration:
ToolDescriptionHuman-in-the-Loop
lsList files and directories-
read_fileRead contents of a file; supports images (.png, .jpg, .jpeg, .gif, .webp) as multimodal content-
write_fileCreate or overwrite a fileRequired1
edit_fileMake targeted edits to existing filesRequired1
globFind files matching a pattern (e.g., **/*.py)-
grepSearch for text patterns across files-
shellExecute shell commands (local mode)Required1
executeExecute commands in remote sandbox (sandbox mode)Required1
web_searchSearch the web using Tavily APIRequired1
fetch_urlFetch and convert web pages to markdownRequired1
taskDelegate work to subagents for parallel executionRequired1
write_todosCreate and manage task lists for complex work-
1: Potentially destructive operations require user approval before execution. To bypass human approval, you can toggle auto-approve or start the deep agent with the auto-approve option:
deepagents --auto-approve
Watch the demo video to see how the Deep Agents CLI works.

​
Quickstart

Set model credentials

Select the model provider(s) you want to use and set credentials as environment variables.
Export as an environment variable:
export OPENAI_API_KEY="your-api-key"

Install and run the CLI

curl -LsSf https://raw.githubusercontent.com/langchain-ai/deepagents/refs/heads/main/libs/cli/scripts/install.sh | bash
Run the CLI
deepagents

Give the agent a task

Create a Python script that prints "Hello, World!"
The agent proposes changes with diffs for your approval before modifying files.
Each model provider requires installing its corresponding LangChain integration package. These are available as optional extras when installing the CLI:
# Install with one provider
uv tool install 'deepagents-cli[anthropic]'

# Install with multiple providers at once
uv tool install 'deepagents-cli[ollama,groq]'

# Add additional packages at a later date
uv tool upgrade deepagents-cli --with langchain-xai
For the full list of supported providers and configuration options, see Custom model providers.On first launch (when no [models].default or [models].recent is set), the CLI auto-selects the first available startup credential in this order: OPENAI_API_KEY, ANTHROPIC_API_KEY, GOOGLE_API_KEY, then GOOGLE_CLOUD_PROJECT (Vertex AI). This startup fallback is intentionally narrow; other supported providers (for example, Groq) are still available via --model, /model, or a saved default model.
deepagents --model anthropic:claude-opus-4-5
Enable web search (optional):
export TAVILY_API_KEY="your-key"
API keys can be set as environment variables or in a .env file.

​
Tracing with LangSmith

Enable LangSmith tracing to see agent operations in a LangSmith project:
  1. Enable LangSmith tracing: 
    export LANGCHAIN_TRACING=true
export LANGCHAIN_API_KEY="your-api-key"
  2. Configure agent tracing for deep agent operations such as tool calls and agent decisions: 
    export DEEPAGENTS_LANGSMITH_PROJECT="my-deep-agent-execution"
  3. If you are building a LangChain app with deep agents, and want to separate agent traces from your app’s traces, also configure LANGSMITH_PROJECT: 
    export LANGSMITH_PROJECT="my-app-calls-to-langchain"
When configured, the CLI displays: 
✓ LangSmith tracing: 'my-project'

​
Configuration

Each agent has its own configuration directory at ~/.deepagents/<agent_name>/. The default agent name is agent. 
# List all configured agents
deepagents list

# Use a specific agent configuration
deepagents --agent mybot

# Use a specific model (provider:model format or auto-detect)
deepagents --model anthropic:claude-sonnet-4-5
deepagents --model gpt-4o

# Auto-approve tool usage (skip human-in-the-loop prompts)
deepagents --auto-approve

# Execute code in a remote sandbox
deepagents --sandbox modal        # or runloop, daytona
deepagents --sandbox-id dbx_123   # reuse existing sandbox
OptionDescription
-a, --agent NAMEUse named agent with separate memory (default: agent)
-M, --model MODELUse a specific model (provider:model)
--model-params JSONExtra kwargs to pass to the model as a JSON string (e.g., '{"temperature": 0.7}')
--default-model [MODEL]Set the default model
--clear-default-modelClear the default model
-r, --resume [ID]Resume a session: -r for most recent, -r <ID> for a specific thread
-m, --message TEXTInitial prompt to auto-submit when the session starts (interactive mode)
-n, --non-interactive TEXTRun a single task non-interactively and exit. Shell is disabled unless --shell-allow-list is set
-q, --quietClean output for piping—only the agent’s response goes to stdout. Requires -n or piped stdin
--no-streamBuffer the full response and write to stdout at once instead of streaming. Requires -n or piped stdin
--auto-approveAuto-approve all tool calls without prompting (disables human-in-the-loop). Toggle with Shift+Tab during an interactive session
--shell-allow-list LISTComma-separated shell commands to auto-approve, 'recommended' for safe defaults, or 'all' to allow any command. Applies to both -n and interactive modes
--sandbox TYPERemote sandbox for code execution: none (default), modal, daytona, runloop, langsmith
--sandbox-id IDReuse an existing sandbox (skips creation and cleanup)
--sandbox-setup PATHPath to setup script to run in sandbox after creation
--mcp-config PATHAdd an explicit MCP config as the highest-precedence source (merged with auto-discovered configs)
--no-mcpDisable all MCP tool loading
--trust-project-mcpTrust project-level MCP configs with stdio servers (skip approval prompt)
-v, --versionDisplay version
-h, --helpShow help
CommandDescription
deepagents helpShow help
deepagents listList all agents
deepagents reset --agent NAMEClear agent memory and reset to default
deepagents reset --agent NAME --target SOURCECopy memory from another agent
deepagents skills list [--project]List all skills (alias: ls)
deepagents skills create NAME [--project]Create a new skill with template SKILL.md
deepagents skills info NAME [--project]Show detailed information about a skill
deepagents skills delete NAME [--project] [-f]Delete a skill and its contents
deepagents threads list [--agent NAME] [--limit N]List sessions (alias: ls). Default limit: 20. -n is a short flag for --limit
deepagents threads delete IDDelete a session

​
Interactive mode

Type naturally as you would in a chat interface. The agent will use its built-in tools, skills, and memory to help you with tasks.
Use these commands within the CLI session:
  • /model - Open the interactive model selector. The selector shows a detail footer with context window size, input modalities, and capabilities for the highlighted model
  • /model <provider:model> - Switch directly to a specific model (e.g., /model anthropic:claude-sonnet-4-5)
  • /model --model-params JSON [provider:model] - Switch with extra model constructor params (e.g., /model --model-params '{"temperature": 0.7}' anthropic:claude-sonnet-4-5). Use Tab to autocomplete the currently highlighted model name
  • /model --default <provider:model> - Set a persistent default model
  • /model --default --clear - Clear the saved default model
  • /remember [context] - Review conversation and update memory and skills. Optionally pass additional context
  • /tokens - Display current context window token usage
  • /clear - Clear conversation history and start a new thread
  • /threads - Browse and resume previous conversation threads. Supports fuzzy search, configurable columns (thread_id, agent_name, messages, created_at, updated_at, git_branch, initial_prompt), sort order toggle, and inline deletion with Ctrl+D. Column visibility and sort preferences persist to config.toml
  • /reload - Refresh runtime configuration (API keys, .env changes, project root, shell allow-list) without restarting the session. Conversation state is preserved. Reports which fields changed (API keys are masked). Does not reload the active model or agent graph
  • /trace - Open the current thread in LangSmith (requires LANGSMITH_API_KEY)
  • /changelog - Open the CLI changelog in your browser
  • /docs - Open the documentation in your browser
  • /feedback - Open the GitHub issues page to file a bug report or feature request
  • /version - Show installed deepagents-cli and SDK versions
  • /help - Show help and available commands
  • /quit - Exit the CLI
Type ! to enter shell mode, then type your command.
git status
npm test
ls -la
General
ShortcutAction
EnterSubmit prompt
Shift+Enter, Ctrl+J, Alt+Enter, or Ctrl+EnterInsert newline
Ctrl+ASelect all text in input
@filenameAuto-complete files and inject content
Shift+Tab or Ctrl+TToggle auto-approve
Ctrl+EExpand/collapse the most recent tool output
EscapeInterrupt current operation
Ctrl+CInterrupt or quit
Ctrl+DExit

​
Non-interactive mode and piping

Use -n to run a single task without launching the interactive UI: 
deepagents -n "Write a Python script that prints hello world"
You can also pipe input via stdin. When input is piped, the CLI automatically runs non-interactively: 
echo "Explain this code" | deepagents
cat error.log | deepagents -n "What's causing this error?"
git diff | deepagents -n "Review these changes"
When piped input is combined with -n or -m, the piped content is prepended to the flag’s value.
The maximum piped input size is 10 MiB.
Use -q for clean output suitable for piping into other commands, and --no-stream to buffer the full response (instead of streaming) before writing to stdout: 
deepagents -n "Generate a .gitignore for Python" -q > .gitignore
deepagents -n "List dependencies" -q --no-stream | sort
By default, shell execution is disabled in non-interactive mode. Use --shell-allow-list to enable it: 
# Allow specific commands (validated against the list)
deepagents -n "Run the tests and fix failures" --shell-allow-list "pytest,git,make"

# Use the curated safe-command list
deepagents -n "Build the project" --shell-allow-list recommended

# Allow any shell command
deepagents -n "Fix the build" --shell-allow-list all
--shell-allow-list all lets the agent execute arbitrary shell commands with no human confirmation. Use with caution.

​
Switch models

You can switch models during a session without restarting the CLI using the /model command, or at launch with the --model flag: 
> /model anthropic:claude-opus-4-5
> /model openai:gpt-4o

deepagents --model openai:gpt-4o
Run /model with no arguments to open an interactive model selector that displays available models grouped by provider. The selector shows a detail footer for the highlighted model with context window size, input modalities (text, image, audio, PDF, video), and capabilities (reasoning, tool calling, structured output). Values overridden by --profile-override or config.toml are marked with a yellow * prefix. You can also pass extra model constructor parameters when switching mid-session using --model-params: 
> /model --model-params '{"temperature": 0.7}' anthropic:claude-sonnet-4-5
> /model --model-params '{"temperature": 0.7}'  # opens selector, applies params to chosen model
These are session-only overrides and take the highest priority, overriding values from config file params. --model-params cannot be combined with --default. For full details on switching models, setting a default, and configuring custom providers, see Model providers. For config.toml reference and lifecycle hooks, see Configuration.

​
Teach your agent project conventions

As you use the agent, it automatically stores information in ~/.deepagents/<agent_name>/memories/ as markdown files using a memory-first protocol:
  1. Research: Searches memory for relevant context before starting tasks
  2. Response: Checks memory when uncertain during execution
  3. Learning: Automatically saves new information for future sessions
The agent organizes its memories by topic with descriptive filenames: 
~/.deepagents/backend-dev/memories/
├── api-conventions.md
├── database-schema.md
└── deployment-process.md
When you teach the agent conventions: 
uvx deepagents-cli --agent backend-dev
> Our API uses snake_case and includes created_at/updated_at timestamps
It remembers for future sessions: 
> Create a /users endpoint
# Applies conventions without prompting

​
Customize your deep agent

There are two primary ways to customize any agent:
  • Memory: Global and project-specific AGENTS.md files which are loaded in full at session start. Use memory for general coding style and preferences.
  • Skills: Global and project-specific context, conventions, guidelines, or instructions. Use skills for context that is only required when performing specific tasks.

​
Provide project or user context

AGENTS.md files provide persistent memory that is always loaded at session start. You can provide global user memory for the agent in ~/.deepagents/<agent_name>/AGENTS.md. This file is always loaded when you start a new deep agent session. The agent may also read its memory files when answering project-specific questions or when you reference past work or patterns. For project-specific memory, you can add context to .deepagents/AGENTS.md in any project’s root folder as long as the project uses git. When you start the CLI from anywhere within the project’s folder, the CLI will find the project’s root folder by checking for a containing .git folder. Both global and project-level AGENTS.md files are loaded together and appended to the system prompt at startup. The agent will update them as you use the agent and provide it with additional information on how it should behave, feedback on its work, or instructions to remember something. The agent will also update its memory if it identifies patterns or preferences from your interactions. If you would like to explicitly prompt your deep agent to update skills and memory based on the current context from the thread use the /remember command which loads a custom instruction to review the context and perform updates. To add more structured project knowledge in additional memory files, you can add them in .deepagents/ and reference them in the AGENTS.md file. You must reference additional files in the AGENTS.md file for the agent to be aware of these files. The additional files will not be read on startup but the agent can reference and update them when needed.
Global AGENTS.md (~/.deepagents/agent/AGENTS.md)
  • Your personality, style, and universal coding preferences
  • General tone and communication style
  • Universal coding preferences (formatting, type hints, etc.)
  • Tool usage patterns that apply everywhere
  • Workflows and methodologies that don’t change per-project
Project AGENTS.md (.deepagents/AGENTS.md in project root)
  • Project-specific context and conventions
  • Project architecture and design patterns
  • Coding conventions specific to this codebase
  • Testing strategies and deployment processes
  • Team guidelines and project structure

​
Use remote sandboxes

Execute code in isolated remote environments for safety and flexibility. Remote sandboxes provide the following benefits:
  • Safety: Protect your local machine from potentially harmful code execution
  • Clean environments: Use specific dependencies or OS configurations without local setup
  • Parallel execution: Run multiple agents simultaneously in isolated environments
  • Long-running tasks: Execute time-intensive operations without blocking your machine
  • Reproducibility: Ensure consistent execution environments across teams
To use a remote sandbox, follow these steps:
  1. Configure your sandbox provider (Runloop, Daytona, or Modal): 
    # Runloop
export RUNLOOP_API_KEY="your-key"

# Daytona
export DAYTONA_API_KEY="your-key"

# Modal
modal setup
  2. Run the CLI with a sandbox: 
    uvx deepagents-cli --sandbox runloop --sandbox-setup ./setup.sh
    The agent runs locally but executes all code operations in the remote sandbox. Optional setup scripts can configure environment variables, clone repositories, and prepare dependencies.
  3. (Optional) Create a setup.sh file to configure your sandbox environment: 
    #!/bin/bash
set -e

# Clone repository using GitHub token
git clone https://x-access-token:${GITHUB_TOKEN}@github.com/username/repo.git $HOME/workspace
cd $HOME/workspace

# Make environment variables persistent
cat >> ~/.bashrc <<'EOF'
export GITHUB_TOKEN="${GITHUB_TOKEN}"
export OPENAI_API_KEY="${OPENAI_API_KEY}"
cd $HOME/workspace
EOF

source ~/.bashrc
    Store secrets in a local .env file for the setup script to access.
Sandboxes isolate code execution, but agents remain vulnerable to prompt injection with untrusted inputs. Use human-in-the-loop approval, short-lived secrets, and trusted setup scripts only.Note that sandbox APIs are evolving rapidly, and we expect more providers to support proxies that help mitigate prompt injection and secrets management concerns.

​
Use MCP tools

Extend the CLI with tools from external MCP (Model Context Protocol) servers. Place a .mcp.json at your project root and the CLI discovers it automatically. See the MCP tools guide for configuration format, auto-discovery, and troubleshooting.

​
Use skills

Skills are reusable agent capabilities that provide specialized workflows and domain knowledge. You can use skills to provide your deep agent with new capabilities and expertise. Deep agent skills follow the Agent Skills standard. Once you have added skills your deep agent will automatically make use of them and update them as you use the agent and provide it with additional information. If you would like to explicitly prompt your deep agent to update skills and memory based on the current context from the thread use the /remember command which loads a custom instruction to review the context and perform updates.

​
Add skills

  1. Start by creating a skill: 
    deepagents skills create test-skill
    This will generate the correct the following files in your ~/.deepagents/<agent_name> folder:
    skills/
└── test-skill
    └── SKILL.md
  2. Open the generated SKILL.md and edit the file to include your instructions.
  3. Optionally add additional scripts or other resources to the test-skill folder. For more information, see Examples.
If you already have a skills folder with containing skills files, you can also copy them directly to the agent’s folder: 
mkdir -p ~/.deepagents/<agent_name>/skills
cp -r examples/skills/web-research ~/.deepagents/<agent_name>/skills/
At startup, the CLI discovers skills from both Deep Agents and shared alias directories: 
`~/.deepagents/<agent_name>/skills/`
`~/.agents/skills/`
`.deepagents/skills/`
`.agents/skills/`.
When duplicate skill names exist, later-precedence directories override earlier ones (see App data). For project-specific skills, the project’s root folder must have a .git folder. When you start the CLI from anywhere within the project’s folder, the CLI will find the project’s root folder by checking for a containing .git folder. For each skill, the CLI reads the name and the description from the SKILL.md file’s frontmatter. As you use the CLI, if a task matches the skill’s description, the agent will read the skill file and follow its instructions.

​
List skills

To see the lists you have installed, run: 
deepagents skills list
To get more information for a specific skill, run: 
deepagents skills info test-skill
Edit this page on GitHub or file an issue.
Connect these docs to Claude, VSCode, and more via MCP for real-time answers.