> ## Documentation Index¶
> Fetch the complete documentation index at: https://code.claude.com/docs/llms.txt¶
> Use this file to discover all available pages before exploring further.¶
¶
# Run Claude Code programmatically¶
¶
> Use the Agent SDK to run Claude Code programmatically from the CLI, Python, or TypeScript.¶
¶
The [Agent SDK](https://platform.claude.com/docs/en/agent-sdk/overview) gives you the same tools, agent loop, and context management that power Claude Code. It's available as a CLI for scripts and CI/CD, or as [Python](https://platform.claude.com/docs/en/agent-sdk/python) and [TypeScript](https://platform.claude.com/docs/en/agent-sdk/typescript) packages for full programmatic control.¶
¶
<Note>¶
The CLI was previously called "headless mode." The `-p` flag and all CLI options work the same way.¶
</Note>¶
¶
To run Claude Code programmatically from the CLI, pass `-p` with your prompt and any [CLI options](/en/cli-reference):¶
¶
```bash theme={null}¶
claude -p "Find and fix the bug in auth.py" --allowedTools "Read,Edit,Bash"¶
```¶
¶
This page covers using the Agent SDK via the CLI (`claude -p`). For the Python and TypeScript SDK packages with structured outputs, tool approval callbacks, and native message objects, see the [full Agent SDK documentation](https://platform.claude.com/docs/en/agent-sdk/overview).¶
¶
## Basic usage¶
¶
Add the `-p` (or `--print`) flag to any `claude` command to run it non-interactively. All [CLI options](/en/cli-reference) work with `-p`, including:¶
¶
* `--continue` for [continuing conversations](#continue-conversations)¶
* `--allowedTools` for [auto-approving tools](#auto-approve-tools)¶
* `--output-format` for [structured output](#get-structured-output)¶
¶
This example asks Claude a question about your codebase and prints the response:¶
¶
```bash theme={null}¶
claude -p "What does the auth module do?"¶
```¶
¶
### Start faster with bare mode¶
¶
Add `--bare` to reduce startup time by skipping auto-discovery of hooks, skills, plugins, MCP servers, auto memory, and CLAUDE.md. Without it, `claude -p` loads the same [context](/en/how-claude-code-works#the-context-window) an interactive session would, including anything configured in the working directory or `~/.claude`.¶
¶
Bare mode is useful for CI and scripts where you need the same result on every machine. A hook in a teammate's `~/.claude` or an MCP server in the project's `.mcp.json` won't run, because bare mode never reads them. Only flags you pass explicitly take effect.¶
¶
This example runs a one-off summarize task in bare mode and pre-approves the Read tool so the call completes without a permission prompt:¶
¶
```bash theme={null}¶
claude --bare -p "Summarize this file" --allowedTools "Read"¶
```¶
¶
In bare mode Claude has access to the Bash, file read, and file edit tools. Pass any context you need with a flag:¶
¶
| To load | Use |¶
| ----------------------- | ------------------------------------------------------- |¶
| System prompt additions | `--append-system-prompt`, `--append-system-prompt-file` |¶
| Settings | `--settings <file-or-json>` |¶
| MCP servers | `--mcp-config <file-or-json>` |¶
| Custom agents | `--agents <json>` |¶
| A plugin directory | `--plugin-dir <path>` |¶
¶
Bare mode skips OAuth and keychain reads. Anthropic authentication must come from `ANTHROPIC_API_KEY` or an `apiKeyHelper` in the JSON passed to `--settings`. Bedrock, Vertex, and Foundry use their usual provider credentials.¶
¶
<Note>¶
`--bare` is the recommended mode for scripted and SDK calls, and will become the default for `-p` in a future release.¶
</Note>¶
¶
## Examples¶
¶
These examples highlight common CLI patterns. For CI and other scripted calls, add [`--bare`](#start-faster-with-bare-mode) so they don't pick up whatever happens to be configured locally.¶
¶
### Get structured output¶
¶
Use `--output-format` to control how responses are returned:¶
¶
* `text` (default): plain text output¶
* `json`: structured JSON with result, session ID, and metadata¶
* `stream-json`: newline-delimited JSON for real-time streaming¶
¶
This example returns a project summary as JSON with session metadata, with the text result in the `result` field:¶
¶
```bash theme={null}¶
claude -p "Summarize this project" --output-format json¶
```¶
¶
To get output conforming to a specific schema, use `--output-format json` with `--json-schema` and a [JSON Schema](https://json-schema.org/) definition. The response includes metadata about the request (session ID, usage, etc.) with the structured output in the `structured_output` field.¶
¶
This example extracts function names and returns them as an array of strings:¶
¶
```bash theme={null}¶
claude -p "Extract the main function names from auth.py" \¶
--output-format json \¶
--json-schema '{"type":"object","properties":{"functions":{"type":"array","items":{"type":"string"}}},"required":["functions"]}'¶
```¶
¶
<Tip>¶
Use a tool like [jq](https://jqlang.github.io/jq/) to parse the response and extract specific fields:¶
¶
```bash theme={null}¶
# Extract the text result¶
claude -p "Summarize this project" --output-format json | jq -r '.result'¶
¶
# Extract structured output¶
claude -p "Extract function names from auth.py" \¶
--output-format json \¶
--json-schema '{"type":"object","properties":{"functions":{"type":"array","items":{"type":"string"}}},"required":["functions"]}' \¶
| jq '.structured_output'¶
```¶
</Tip>¶
¶
### Stream responses¶
¶
Use `--output-format stream-json` with `--verbose` and `--include-partial-messages` to receive tokens as they're generated. Each line is a JSON object representing an event:¶
¶
```bash theme={null}¶
claude -p "Explain recursion" --output-format stream-json --verbose --include-partial-messages¶
```¶
¶
The following example uses [jq](https://jqlang.github.io/jq/) to filter for text deltas and display just the streaming text. The `-r` flag outputs raw strings (no quotes) and `-j` joins without newlines so tokens stream continuously:¶
¶
```bash theme={null}¶
claude -p "Write a poem" --output-format stream-json --verbose --include-partial-messages | \¶
jq -rj 'select(.type == "stream_event" and .event.delta.type? == "text_delta") | .event.delta.text'¶
```¶
¶
When an API request fails with a retryable error, Claude Code emits a `system/api_retry` event before retrying. You can use this to surface retry progress or implement custom backoff logic.¶
¶
| Field | Type | Description |¶
| ---------------- | --------------- | -------------------------------------------------------------------------------------------------------------------------------------------- |¶
| `type` | `"system"` | message type |¶
| `subtype` | `"api_retry"` | identifies this as a retry event |¶
| `attempt` | integer | current attempt number, starting at 1 |¶
| `max_retries` | integer | total retries permitted |¶
| `retry_delay_ms` | integer | milliseconds until the next attempt |¶
| `error_status` | integer or null | HTTP status code, or `null` for connection errors with no HTTP response |¶
| `error` | string | error category: `authentication_failed`, `billing_error`, `rate_limit`, `invalid_request`, `server_error`, `max_output_tokens`, or `unknown` |¶
| `uuid` | string | unique event identifier |¶
| `session_id` | string | session the event belongs to |¶
¶
For programmatic streaming with callbacks and message objects, see [Stream responses in real-time](https://platform.claude.com/docs/en/agent-sdk/streaming-output) in the Agent SDK documentation.¶
¶
### Auto-approve tools¶
¶
Use `--allowedTools` to let Claude use certain tools without prompting. This example runs a test suite and fixes failures, allowing Claude to execute Bash commands and read/edit files without asking for permission:¶
¶
```bash theme={null}¶
claude -p "Run the test suite and fix any failures" \¶
--allowedTools "Bash,Read,Edit"¶
```¶
¶
To set a baseline for the whole session instead of listing individual tools, pass a [permission mode](/en/permission-modes). `dontAsk` denies anything not in your `permissions.allow` rules, which is useful for locked-down CI runs. `acceptEdits` lets Claude write files without prompting, but shell commands and network requests still need an `--allowedTools` entry or a `permissions.allow` rule, otherwise the run aborts when one is attempted:¶
¶
```bash theme={null}¶
claude -p "Apply the lint fixes" --permission-mode acceptEdits¶
```¶
¶
### Create a commit¶
¶
This example reviews staged changes and creates a commit with an appropriate message:¶
¶
```bash theme={null}¶
claude -p "Look at my staged changes and create an appropriate commit" \¶
--allowedTools "Bash(git diff *),Bash(git log *),Bash(git status *),Bash(git commit *)"¶
```¶
¶
The `--allowedTools` flag uses [permission rule syntax](/en/settings#permission-rule-syntax). The trailing ` *` enables prefix matching, so `Bash(git diff *)` allows any command starting with `git diff`. The space before `*` is important: without it, `Bash(git diff*)` would also match `git diff-index`.¶
¶
<Note>¶
User-invoked [skills](/en/skills) like `/commit` and [built-in commands](/en/commands) are only available in interactive mode. In `-p` mode, describe the task you want to accomplish instead.¶
</Note>¶
¶
### Customize the system prompt¶
¶
Use `--append-system-prompt` to add instructions while keeping Claude Code's default behavior. This example pipes a PR diff to Claude and instructs it to review for security vulnerabilities:¶
¶
```bash theme={null}¶
gh pr diff "$1" | claude -p \¶
--append-system-prompt "You are a security engineer. Review for vulnerabilities." \¶
--output-format json¶
```¶
¶
See [system prompt flags](/en/cli-reference#system-prompt-flags) for more options including `--system-prompt` to fully replace the default prompt.¶
¶
### Continue conversations¶
¶
Use `--continue` to continue the most recent conversation, or `--resume` with a session ID to continue a specific conversation. This example runs a review, then sends follow-up prompts:¶
¶
```bash theme={null}¶
# First request¶
claude -p "Review this codebase for performance issues"¶
¶
# Continue the most recent conversation¶
claude -p "Now focus on the database queries" --continue¶
claude -p "Generate a summary of all issues found" --continue¶
```¶
¶
If you're running multiple conversations, capture the session ID to resume a specific one:¶
¶
```bash theme={null}¶
session_id=$(claude -p "Start a review" --output-format json | jq -r '.session_id')¶
claude -p "Continue that review" --resume "$session_id"¶
```¶
¶
## Next steps¶
¶
* [Agent SDK quickstart](https://platform.claude.com/docs/en/agent-sdk/quickstart): build your first agent with Python or TypeScript¶
* [CLI reference](/en/cli-reference): all CLI flags and options¶
* [GitHub Actions](/en/github-actions): use the Agent SDK in GitHub workflows¶
* [GitLab CI/CD](/en/gitlab-ci-cd): use the Agent SDK in GitLab pipelines¶