Claude Help Center

https://support.claude.com/en/
Admin
The Claude Help Center provides a comprehensive collection of articles and resources to assist users...
Tokens:521,836
Snippets:5,095
Trust Score:10
Update:2 weeks ago
Show doc for...
Tokens:
Raw
### Install and run Claude Code

Source: https://support.claude.com/en/articles/14555877-claude-code-communications-kit

Install Claude Code and start it in your repository. Run /init once after starting to generate a CLAUDE.md file with build commands and conventions.

```bash
curl -fsSL https://claude.ai/install.sh | bash  
cd <your-repo>  
claude
```

--------------------------------

### Install QEMU packages on x64

Source: https://support.claude.com/en/articles/10065433-installing-claude-desktop

Installs QEMU and supporting packages required for Claude Cowork on Linux x64 when the .deb file was installed directly.

```bash
sudo apt install qemu-system-x86 ovmf virtiofsd
```

--------------------------------

### Install QEMU packages on arm64

Source: https://support.claude.com/en/articles/10065433-installing-claude-desktop

Installs QEMU and supporting packages for Claude Cowork on Linux arm64. Note that this command omits the 'sudo apt install' prefix as shown in the source.

```bash
qemu-system-arm qemu-efi-aarch64 virtiofsd
```

--------------------------------

### Install Claude Desktop from .deb file

Source: https://support.claude.com/en/articles/10065433-installing-claude-desktop

Install Claude Desktop from a downloaded .deb file when apt installation is not possible. Replace the filename with the correct architecture (x64 or arm64) if needed. Note that this method does not opt into regular package updates.

```bash
sudo apt install ~/Downloads/claude-desktop_amd64.deb
```

--------------------------------

### Set HTTPS proxy for installer

Source: https://support.claude.com/en/articles/14552646-troubleshoot-claude-code-installation-and-authentication

Set your corporate proxy before re-running the installer when the download host storage.googleapis.com is blocked.

```bash
export HTTPS_PROXY=http://proxy.example.com:port
```

--------------------------------

### Open a project

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Navigate to your project directory and start Claude Code from there.

```bash
cd path/to/your/project claude
```

--------------------------------

### Verify Claude Code installation

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Run this command to verify that Claude Code was installed correctly.

```bash
claude --version
```

--------------------------------

### Install Claude Code (curl, PowerShell, Homebrew, WinGet)

Source: https://support.claude.com/en/articles/14554922-claude-code-user-faq

Install Claude Code on macOS/Linux via curl, Windows via PowerShell, or using Homebrew or WinGet package managers. After installation, run 'claude' from any repository.

```bash
curl -fsSL https://claude.ai/install.sh | bash
```

```powershell
irm https://claude.ai/install.ps1 | iex
```

```bash
brew install --cask claude-code
```

```powershell
winget install Anthropic.ClaudeCode
```

--------------------------------

### Get a spend limit - example curl request

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example request for getting a single spend limit by ID. Uses the x-api-key header with the ANTHROPIC_ADMIN_KEY environment variable. The spend_limit_id path parameter is prefixed with spl_.

```bash
curl "https://api.anthropic.com/v1/organizations/spend_limits/spl_01AbCdEfGhIjKlMnOpQrSt" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

--------------------------------

### Install Claude Desktop via apt

Source: https://support.claude.com/en/articles/10065433-installing-claude-desktop

Install Claude Desktop on Ubuntu or Debian from Anthropic's apt repository. Run these commands in a terminal to add the repository key, configure the apt source, and install the package. Updates arrive through the system's regular package updates.

```bash
sudo curl -fsSLo /usr/share/keyrings/claude-desktop-archive-keyring.asc https://downloads.claude.ai/claude-desktop/key.asc  
echo "deb [signed-by=/usr/share/keyrings/claude-desktop-archive-keyring.asc] https://downloads.claude.ai/claude-desktop/apt/stable stable main" | sudo tee /etc/apt/sources.list.d/claude-desktop.list  
sudo apt update && sudo apt install claude-desktop
```

--------------------------------

### Start Claude Code with --model flag for Haiku 4.5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Haiku 4.5 model for the current session only.

```bash
claude --model claude-haiku-4-5-20251001
```

--------------------------------

### Onboard to a new repository: /init and codebase tour

Source: https://support.claude.com/en/articles/14553517-claude-code-common-developer-use-cases

Use this prompt when working in a repository for the first time and needing a structured overview of its architecture, build commands, and conventions. `/init` scans the project and writes a `CLAUDE.md` file summarizing build commands, architecture, and conventions; the follow-up tour prompt produces a guided walkthrough. Project context remains loaded for the rest of the session.

```text
> /init
> give me a tour of this codebase — where's the entry point, how is it structured, what should I read first
```

--------------------------------

### Get oriented

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Ask Claude to summarize the codebase and identify the entry point.

```text
Give me a 5-bullet summary of what this codebase does and where the entry point is.
```

--------------------------------

### SpendLimit object example

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example of a SpendLimit object at the user scope. The amount is in minor units and null means unlimited. The id is prefixed with spl_.

```json
{  
  "type": "spend_limit",  
  "id": "spl_01AbCdEfGhIjKlMnOpQrSt",  
  "created_at": "2026-05-01T12:00:00Z",  
  "updated_at": "2026-05-03T09:14:11Z",  
  "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },  
  "amount": "50000",  
  "currency": "USD",  
  "period": "monthly"  
}
```

--------------------------------

### Start Claude Code with --model flag for Fable 5.1

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Fable 5.1 model for the current session only.

```bash
claude --model claude-fable-5-1
```

--------------------------------

### Start Claude Code with --model flag for Opus 4.5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Opus 4.5 model for the current session only.

```bash
claude --model claude-opus-4-5-20251101
```

--------------------------------

### Example skill.md for Brand Guidelines Skill

Source: https://support.claude.com/en/articles/12512198-how-to-create-custom-skills

Example skill.md file for a brand guidelines skill. It includes required YAML frontmatter with name and description, followed by a markdown body with sections for brand colors, typography, logo usage, and when to apply. The description is critical for Claude to determine when to invoke the skill.

```markdown
## Metadata  
name: Brand Guidelines  
description: Apply Acme Corp brand guidelines to all presentations and documents  
  
## Overview  
This skill provides Acme Corp's official brand guidelines for creating consistent, professional materials. When creating presentations, documents, or marketing materials, apply these standards to ensure all outputs match Acme's visual identity. Claude should reference these guidelines whenever creating external-facing materials or documents that represent Acme Corp.  
  
## Brand Colors  
  
Our official brand colors are:  
- Primary: #FF6B35 (Coral)  
- Secondary: #004E89 (Navy Blue)  
- Accent: #F7B801 (Gold)  
- Neutral: #2E2E2E (Charcoal)  
  
## Typography  
  
Headers: Montserrat Bold  
Body text: Open Sans Regular  
Size guidelines:  
- H1: 32pt  
- H2: 24pt  
- Body: 11pt  
  
## Logo Usage  
  
Always use the full-color logo on light backgrounds. Use the white logo on dark backgrounds. Maintain minimum spacing of 0.5 inches around the logo.  
  
## When to Apply  
  
Apply these guidelines whenever creating:  
- PowerPoint presentations  
- Word documents for external sharing  
- Marketing materials  
- Reports for clients  
  
## Resources  
  
See the resources folder for logo files and font downloads.
```

--------------------------------

### Start Claude Code with --model flag for Opus 5.5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Opus 5.5 model for the current session only.

```bash
claude --model claude-opus-5-5
```

--------------------------------

### Start Claude Code with --model flag for Sonnet 4.5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Sonnet 4.5 model for the current session only.

```bash
claude --model claude-sonnet-4-5-20250929
```

--------------------------------

### Start Claude Code with --model flag for Fable 5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Fable 5 model for the current session only.

```bash
claude --model claude-fable-5
```

--------------------------------

### Start Claude Code with --model flag for Opus 4.7

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Opus 4.7 model for the current session only.

```bash
claude --model claude-opus-4-7
```

--------------------------------

### Start Claude Code with --model flag for Opus 4.6

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Opus 4.6 model for the current session only.

```bash
claude --model claude-opus-4-6
```

--------------------------------

### Get a spend limit increase request with curl

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example request to get a specific spend limit increase request by its ID. Uses the x-api-key header with the ANTHROPIC_ADMIN_KEY environment variable.

```bash
curl "https://api.anthropic.com/v1/organizations/spend_limit_increase_requests/slir_01AbCdEfGhIjKlMnOpQrSt" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

--------------------------------

### Start Claude Code with --model flag for Opus 4.8

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Opus 4.8 model for the current session only.

```bash
claude --model claude-opus-4-8
```

--------------------------------

### Start a Claude Code session with claude:// URL

Source: https://support.claude.com/en/articles/14729294-open-claude-desktop-with-a-link

Opens Claude Code in Claude Desktop with the composer prefilled. Use this for "Open in Claude Code" buttons in your own tools. The q parameter accepts prompt as an alias. The folder parameter is an absolute path to use as the working directory; Claude Desktop asks you to confirm the folder before it's used, and only the first folder value is applied today. The file parameter is accepted but not currently supported. Any folder supplied through a link is treated as untrusted, and Claude Desktop always shows a confirmation dialog before adopting it as the working directory, even if you've trusted that folder before.

```text
claude://code/new  
claude://code/new?q=Fix%20the%20failing%20test&folder=%2FUsers%2Fme%2Frepo
```

--------------------------------

### Start Claude Code with --model flag for Opus 5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Opus 5 model for the current session only.

```bash
claude --model claude-opus-5
```

--------------------------------

### Start a Claude Cowork session with claude:// URL

Source: https://support.claude.com/en/articles/14729294-open-claude-desktop-with-a-link

Opens a new Cowork session with the composer prefilled. Use this for "Open in Claude Cowork" buttons in your own tools. The q parameter prefills the composer. The folder parameter is an absolute path to a folder to attach; repeat the parameter to attach more than one. The file parameter is an absolute path to a file to attach; repeat the parameter to attach more than one. Claude Desktop asks you to confirm each folder before it's used. Any folder supplied through a link is treated as untrusted, and Claude Desktop always shows a confirmation dialog before adopting it as the working directory, even if you've trusted that folder before.

```text
claude://cowork/new  
claude://cowork/new?q=Draft%20the%20Q2%20update&folder=%2FUsers%2Fme%2Fdocs&file=%2FUsers%2Fme  
%2Fdocs%2Fnotes.md
```

--------------------------------

### Test deep link on Windows with start command

Source: https://support.claude.com/en/articles/14729294-open-claude-desktop-with-a-link

Test a deep link on Windows by running this command in Command Prompt.

```cmd
start "" "claude://claude.ai/new?q=Hello"
```

--------------------------------

### Start Claude Code with --model flag for Sonnet 4.6

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Sonnet 4.6 model for the current session only.

```bash
claude --model claude-sonnet-4-6
```

--------------------------------

### Install Claude Code via Homebrew

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Use this command to install Claude Code using Homebrew on macOS.

```bash
brew install --cask claude-code
```

--------------------------------

### Start Claude Code with --model flag for Sonnet 5.5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Sonnet 5.5 model for the current session only.

```bash
claude --model claude-sonnet-5-5
```

--------------------------------

### Install Claude Code on macOS/Linux/WSL

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Use this command to install Claude Code on macOS, Linux, or Windows Subsystem for Linux (WSL).

```bash
curl -fsSL https://claude.ai/install.sh
```

--------------------------------

### List effective spend limits - example response

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example response for the effective spend limits endpoint. Shows a SpendSummary entry with scope, amount, currency, period, source, spend_limit_id, and period_to_date_spend fields, plus a next_page cursor.

```json
{
  "data": [
    {
      "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },
      "amount": "50000",
      "currency": "USD",
      "period": "monthly",
      "source": { "type": "seat_tier", "seat_tier": "enterprise_standard" },
      "spend_limit_id": "spl_01XyZaBcDeFgHiJkLmNoPq",
      "period_to_date_spend": "31402.5"
    }
  ],
  "next_page": "page_..."
}
```

--------------------------------

### Install Claude Code on Windows (PowerShell)

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Use this command to install Claude Code on Windows using PowerShell.

```powershell
irm https://claude.ai/install.ps1
```

--------------------------------

### Start Claude Code with claude command

Source: https://support.claude.com/en/articles/11940350

After applying the changes and reopening the terminal, run this command to start Claude Code with the new default model.

```bash
claude
```

--------------------------------

### /schedule command for cloud jobs

Source: https://support.claude.com/en/articles/14554000-claude-code-power-user-tips

Example of using /schedule to create a daily cloud job that reviews PRs and updates docs via Slack MCP. Unlike /loop, scheduled jobs run in the cloud and continue even when the laptop is closed.

```bash
> /schedule a daily job that looks at all PRs shipped since   yesterday and updates our docs based on the changes. Use   the Slack MCP to message #docs-update with the changes
```

--------------------------------

### Start a new chat with claude:// URL

Source: https://support.claude.com/en/articles/14729294-open-claude-desktop-with-a-link

Opens a new chat in Claude Desktop. If you include the q parameter, the prompt field is prefilled with that text so you can review and send it.

```text
claude://claude.ai/new  
claude://claude.ai/new?q=Summarize%20this%20week%27s%20release%20notes
```

--------------------------------

### /loop examples for local recurring tasks

Source: https://support.claude.com/en/articles/14554000-claude-code-power-user-tips

Shows three /loop commands for scheduling local recurring tasks: /babysit for auto-addressing reviews and rebasing, /slack-feedback for putting up PRs, and /pr-pruner for closing stale PRs. Each runs for a specified duration (5m, 30m, 1h).

```bash
/loop 5m /babysit         # auto-address review, rebase, shepherd PRs
/loop 30m /slack-feedback # auto put up PRs for Slack feedback
/loop 1h /pr-pruner       # close out stale PRs
```

--------------------------------

### Start Claude Code with --model flag for Sonnet 5

Source: https://support.claude.com/en/articles/11940350-claude-code-model-configuration

Starts Claude Code with the Sonnet 5 model for the current session only.

```bash
claude --model claude-sonnet-5
```

--------------------------------

### Install Claude Code via npm

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Use this command to install Claude Code globally via npm (requires Node 18+). Do not use sudo.

```bash
npm install -g @anthropic-ai/claude-code
```

--------------------------------

### Set spend limit response

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example response for setting a spend limit, showing the SpendLimit object with type, id, timestamps, scope, amount, currency, and period.

```json
{
  "type": "spend_limit",
  "id": "spl_01RsTuVwXyZaBcDeFgHiJk",
  "created_at": "2026-05-11T10:02:44Z",
  "updated_at": "2026-05-11T10:02:44Z",
  "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },
  "amount": "75000",
  "currency": "USD",
  "period": "monthly"
}
```

--------------------------------

### Install Claude MSIX for single user

Source: https://support.claude.com/en/articles/12622703-deploy-claude-desktop-for-windows

Registers the Claude MSIX package for the current user only. Requires the MSIX file to be present at the specified path.

```powershell
powershell  
Add-AppxPackage -Path "Claude.msix"
```

--------------------------------

### List effective spend limits - example curl request

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example request for listing effective spend limits. Uses the x-api-key header with the ANTHROPIC_ADMIN_KEY environment variable. The limit query parameter is set to 20.

```bash
curl "https://api.anthropic.com/v1/organizations/spend_limits/effective?limit=20" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

--------------------------------

### SpendSummary object example

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example of a SpendSummary object, a computed per-member report row. It is not an addressable resource and has no id. The source field indicates where the amount was resolved from in the hierarchy.

```json
{  
  "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },  
  "amount": "50000",  
  "currency": "USD",  
  "period": "monthly",  
  "source": { "type": "seat_tier", "seat_tier": "enterprise_standard" },  
  "spend_limit_id": "spl_01XyZaBcDeFgHiJkLmNoPq",  
  "period_to_date_spend": "31402.5"  
}
```

--------------------------------

### Approve spend limit increase request curl example

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example curl request to approve a spend limit increase request. Sets a new per-user limit of 75000 minor units for the monthly period and suppresses the notification email.

```bash
curl -X POST "https://api.anthropic.com/v1/organizations/spend_limit_increase_requests/slir_01AbCdEfGhIjKlMnOpQrSt/approve" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY" \
  -H "content-type: application/json" \
  -d '{"amount": "75000", "suppress_notification": true}'
```

--------------------------------

### Onboard to a repo

Source: https://support.claude.com/en/articles/14553517-claude-code-common-developer-use-cases

Sample prompt for onboarding to a repository using /init.

```text
/init then “give me a tour”
```

--------------------------------

### Fix npm EACCES with native installer or npm prefix

Source: https://support.claude.com/en/articles/14552646-troubleshoot-claude-code-installation-and-authentication

Fix npm EACCES permission errors by avoiding sudo. Use the native installer or change npm's global prefix to a user-owned directory and add that bin directory to your PATH.

```bash
curl -fsSL https://claude.ai/install.sh | bash
```

```bash
npm config set prefix ~/.npm-global
```

--------------------------------

### Check and update Node version for Claude Code

Source: https://support.claude.com/en/articles/14552646-troubleshoot-claude-code-installation-and-authentication

Check your Node version and install a current LTS version with nvm. Claude Code requires Node 18 or later; the native installer bundles its own runtime and avoids this issue.

```bash
node -v
```

```bash
nvm install --lts
```

--------------------------------

### Get a spend limit (GET /v1/organizations/spend_limits/{spend_limit_id})

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Returns a single spend limit by ID. Use this to inspect the row that a SpendSummary.spend_limit_id or a POST response referenced. Requires the read:spend_limits scope.

```http
GET /v1/organizations/spend_limits/{spend_limit_id}
```

--------------------------------

### Change model with /model command

Source: https://support.claude.com/en/articles/11940350

Start Claude Code, then use the /model command to open an interactive menu for selecting a model. The change takes effect immediately without restarting. Use /status to check the current model.

```bash
claude
```

```bash
/model
```

```bash
/status
```

--------------------------------

### List spend limit increase requests response

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example response for listing spend limit increase requests. Shows a pending request with actor and spend summary details, and a next_page cursor of null.

```json
{
  "data": [
    {
      "type": "spend_limit_increase_request",
      "id": "slir_01AbCdEfGhIjKlMnOpQrSt",
      "created_at": "2026-05-04T16:22:09Z",
      "status": "pending",
      "resolved_at": null,
      "resolved_by": null,
      "actor": {
        "type": "user_actor",
        "user_id": "user_01AbCdEfGh",
        "name": "Jane Smith",
        "email_address": "jane@example.com"
      },
      "spend_summary": {
        "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },
        "amount": "50000",
        "currency": "USD",
        "period": "monthly",
        "source": { "type": "seat_tier", "seat_tier": "enterprise_standard" },
        "spend_limit_id": "spl_01XyZaBcDeFgHiJkLmNoPq",
        "period_to_date_spend": "48900"
      }
    }
  ],
  "next_page": null
}
```

--------------------------------

### Deny spend limit increase request curl example

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example curl request to deny a spend limit increase request. Suppresses the notification email to the member.

```bash
curl -X POST "https://api.anthropic.com/v1/organizations/spend_limit_increase_requests/slir_01AbCdEfGhIjKlMnOpQrSt/deny" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY" \
  -H "content-type: application/json" \
  -d '{"suppress_notification": true}'
```

--------------------------------

### Update Claude Code via WinGet

Source: https://support.claude.com/en/articles/14554922-claude-code-user-faq

Update Claude Code when installed via WinGet, which does not auto-update. Run periodically to get the latest version.

```powershell
winget upgrade Anthropic.ClaudeCode
```

--------------------------------

### Define a read-only agent in .claude/agents/ReadOnly.md

Source: https://support.claude.com/en/articles/14554000-claude-code-power-user-tips

Example of a custom agent definition with frontmatter specifying name, description, color, and tools.

```markdown
# .claude/agents/ReadOnly.md
---
name: ReadOnly
description: Read-only agent restricted to the Read tool only
color: blue
tools: Read
---

You are a read-only agent that cannot edit files or run bash.
```

--------------------------------

### Enable VirtualMachinePlatform for Cowork

Source: https://support.claude.com/en/articles/12622703-deploy-claude-desktop-for-windows

Installs the Virtual Machine Platform feature required for Claude Cowork on Windows. The -NoRestart flag suppresses the automatic restart for silent deployment; the feature takes effect after the machine restarts.

```powershell
powershell  
Enable-WindowsOptionalFeature -Online -FeatureName VirtualMachinePlatform -All -NoRestart
```

--------------------------------

### Update Claude Code via Homebrew

Source: https://support.claude.com/en/articles/14554922-claude-code-user-faq

Update Claude Code when installed via Homebrew, which does not auto-update. Run periodically to get the latest version.

```bash
brew upgrade claude-code
```

--------------------------------

### Update Claude Code (claude update)

Source: https://support.claude.com/en/articles/14554922-claude-code-user-faq

Force an immediate update of Claude Code for native installs, which otherwise auto-update in the background. Homebrew and WinGet installs do not auto-update; use 'brew upgrade claude-code' or 'winget upgrade Anthropic.ClaudeCode' instead.

```bash
claude update
```

--------------------------------

### Securely load API key with python-dotenv

Source: https://support.claude.com/en/articles/9767949-api-key-best-practices-keeping-your-keys-safe-and-secure

Steps to securely store and load the Claude API key using a local .env file and the python-dotenv package. Create a .env file with the key, install python-dotenv, then load the key in your Python script with load_dotenv() and os.getenv().

```plaintext
ANTHROPIC_API_KEY=your-api-key-here
```

```bash
pip install python-dotenv
```

```python
from dotenv import load_dotenv  
  
import os  
  
load_dotenv()  
my_api_key = os.getenv("ANTHROPIC_API_KEY")
```

--------------------------------

### Run claude -p with --bare for faster startup

Source: https://support.claude.com/en/articles/14554000-claude-code-power-user-tips

Use for non-interactive Claude Code invocations to skip searching for local CLAUDE.md files, settings, and MCPs, resulting in roughly 10× faster startup. Specify configuration explicitly via flags like --system-prompt, --mcp-config, or --settings when needed. The default will flip to --bare in a future version.

```bash
claude -p "summarize this codebase" \  
    --output-format=stream-json \  
    --verbose \  
    --bare
```

--------------------------------

### GET /v1/organizations/spend_limit_increase_requests

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Lists all spend limit increase requests submitted by members.

```APIDOC
## GET /v1/organizations/spend_limit_increase_requests

### Description
List members' requests for a higher spend limit.

### Method
GET

### Endpoint
/v1/organizations/spend_limit_increase_requests

### Response
#### Success Response (200)
- **requests** (array) - List of spend limit increase requests.

#### Response Example
{
  "requests": []
}
```

--------------------------------

### Approve spend limit increase request response

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example response for approving a spend limit increase request. Shows the request in status approved with an additional spend_limit field containing the written SpendLimit.

```json
{
  "type": "spend_limit_increase_request",
  "id": "slir_01AbCdEfGhIjKlMnOpQrSt",
  "created_at": "2026-05-04T16:22:09Z",
  "status": "approved",
  "resolved_at": "2026-05-11T10:05:02Z",
  "resolved_by": {
    "type": "scoped_api_key_actor",
    "scoped_api_key_id": "apikey_01ZyXwVuTsRqPoNmLkJiHg"
  },
  "actor": {
    "type": "user_actor",
    "user_id": "user_01AbCdEfGh",
    "name": "Jane Smith",
    "email_address": "jane@example.com"
  },
  "spend_summary": null,
  "spend_limit": {
    "type": "spend_limit",
    "id": "spl_01RsTuVwXyZaBcDeFgHiJk",
    "created_at": "2026-05-11T10:05:02Z",
    "updated_at": "2026-05-11T10:05:02Z",
    "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },
    "amount": "75000",
    "currency": "USD",
    "period": "monthly"
  }
}
```

--------------------------------

### Set spend limit curl request

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example request to set a per-user spend limit. Uses the admin API key and a JSON body with scope and amount.

```bash
curl -X POST "https://api.anthropic.com/v1/organizations/spend_limits" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY" \
  -H "content-type: application/json" \
  -d '{"scope": {"type": "user", "user_id": "user_01AbCdEfGh"}, "amount": "75000"}'
```

--------------------------------

### Create a skill

Source: https://support.claude.com/en/articles/14553517-claude-code-common-developer-use-cases

Sample prompt for creating a skill. Replace [name] and [steps] with the skill name and steps.

```text
make me a /[name] skill that [steps]
```

--------------------------------

### GET /v1/organizations/spend_limits/effective

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Retrieves the effective spend limits for the organization. This endpoint is paginated with an opaque cursor.

```APIDOC
## GET /v1/organizations/spend_limits/effective

### Description
Retrieves the effective spend limits for the organization. This endpoint is paginated with an opaque cursor.

### Method
GET

### Endpoint
/v1/organizations/spend_limits/effective

### Parameters
#### Query Parameters
- **limit** (integer) - Optional - Maximum number of rows to return in the response.
- **page** (string) - Optional - Opaque cursor for pagination. Pass the `next_page` cursor from the previous response unchanged.
- **user_ids[]** (array of strings) - Optional - Filter results by user IDs. Changing this parameter mid-sequence invalidates the cursor.
- **status[]** (array of strings) - Optional - Filter results by status. Changing this parameter mid-sequence invalidates the cursor.
- **actor_ids[]** (array of strings) - Optional - Filter results by actor IDs. Changing this parameter mid-sequence invalidates the cursor.

### Response
#### Success Response (200)
- **rows** (array) - List of effective spend limit records.
- **next_page** (string or null) - Opaque cursor for the next page. `null` indicates no more pages.

#### Response Example
{
  "rows": [],
  "next_page": null
}

### Error Handling
- **429 Too Many Requests** - Rate limit exceeded (60 requests per minute per organization).
- **400 Bad Request** - Cursor does not match current query parameters.
```

--------------------------------

### GET /v1/organizations/spend_limit_increase_requests

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Lists spend limit increase requests, most recent first. Requires scope: `read:spend_limits`.

```APIDOC
## GET /v1/organizations/spend_limit_increase_requests

### Description
Lists spend limit increase requests, most recent first. Requests whose requester is no longer a member of the organization are excluded.

### Method
GET

### Endpoint
/v1/organizations/spend_limit_increase_requests

### Parameters
#### Query Parameters
- **status[]** (one or more of `pending`, `approved`, `denied`) - Optional - Filter by status. Repeat the parameter for multiple values. Default: all.
- **actor_ids[]** (string) - Optional - Filter by requester. Accepts `user_...` IDs. Default: all.
- **limit** (integer 1–1000) - Optional - Rows per page. Default: 20.
- **page** (opaque cursor string) - Optional - The `next_page` value from a previous response.

### Request Example
```
curl "https://api.anthropic.com/v1/organizations/spend_limit_increase_requests?status[]=pending&limit=50" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

### Response
#### Success Response (200)
- **data** (array of SpendLimitIncreaseRequest) - Ordered by `created_at` descending.
- **next_page** (string or null) - Opaque cursor for the next page; `null` when no more pages.

#### Response Example
```
{
  "data": [
    {
      "type": "spend_limit_increase_request",
      "id": "slir_01AbCdEfGhIjKlMnOpQrSt",
      "created_at": "2026-05-04T16:22:09Z",
      "status": "pending",
      "resolved_at": null,
      "resolved_by": null,
      "actor": {
        "type": "user_actor",
        "user_id": "user_01AbCdEfGh",
        "name": "Jane Smith",
        "email_address": "jane@example.com"
      },
      "spend_summary": {
        "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },
        "amount": "50000",
        "currency": "USD",
        "period": "monthly",
        "source": { "type": "seat_tier", "seat_tier": "enterprise_standard" },
        "spend_limit_id": "spl_01XyZaBcDeFgHiJkLmNoPq",
        "period_to_date_spend": "48900"
      }
    }
  ],
  "next_page": null
}
```

### Error Handling
#### Validations
- **Condition**: `actor_ids[]` entry malformed - Status: 400 - Message: `actor_ids[]: invalid tagged user ID`
- **Condition**: `limit` outside 1–1000 - Status: 400 - Message: (none)
- **Condition**: `page` cursor malformed - Status: 400 - Message: `invalid page cursor format` or `invalid page cursor`
- **Condition**: `page` cursor does not match current `status[]` or `actor_ids[]` - Status: 400 - Message: `page cursor does not match current query parameters`
- **Condition**: `page` cursor from a different API version - Status: 400 - Message: `page cursor was issued by a different API version; restart pagination`
- **Condition**: Organization is not on an Enterprise plan - Status: 400 - Message: `this endpoint is not supported for this organization type`
- **Condition**: Usage credit billing not enabled - Status: 400 - Message: `overage billing is not enabled for this organization`
```

--------------------------------

### Understand unfamiliar code

Source: https://support.claude.com/en/articles/14553517-claude-code-common-developer-use-cases

Sample prompt for understanding unfamiliar code. Replace [module] with the module name.

```text
walk me through how [module] works
```

--------------------------------

### List spend limit increase requests with curl

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Example request to list pending spend limit increase requests with a limit of 50. Uses the x-api-key header with the ANTHROPIC_ADMIN_KEY environment variable.

```bash
curl "https://api.anthropic.com/v1/organizations/spend_limit_increase_requests?status[]=pending&limit=50" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

--------------------------------

### Add Claude Code to PATH

Source: https://support.claude.com/en/articles/14554922-claude-code-user-faq

Add the Claude Code binary directory to your PATH in ~/.zshrc or ~/.bashrc to fix 'claude: command not found'. The native installer places the binary at ~/.local/bin/claude (Windows: %USERPROFILE%\.local\bin).

```bash
export PATH="$PATH:$HOME/.local/bin"
```

--------------------------------

### GET /v1/organizations/spend_limits/effective

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Lists effective spend limits for all current members of the organization, with optional filtering and pagination.

```APIDOC
## GET /v1/organizations/spend_limits/effective

### Description
Returns every current member of the organization with their resolved effective limit and period-to-date spend. Members without a per-user override appear with `source.type` of `seat_tier`, `rbac_group`, or `organization`. Ex-members aren’t listed.

### Method
GET

### Endpoint
/v1/organizations/spend_limits/effective

### Parameters
#### Query Parameters
- **user_ids[]** (string, max 100 entries) - Optional - Narrow to specific members. Accepts `user_...` IDs. An entry that isn’t a current member is silently omitted from `data`.
- **limit** (integer 1–1000) - Optional - Rows per page. Default: 20.
- **page** (opaque cursor string) - Optional - The `next_page` value from a previous response.

### Request Example
```
curl "https://api.anthropic.com/v1/organizations/spend_limits/effective?limit=20" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

### Response
#### Success Response (200)
- **data** (array of SpendSummary) - One entry per member, ordered by when the member joined the organization (newest first).
- **next_page** (string or null) - Opaque cursor for the next page; `null` when no more pages.

#### Response Example
```
{
  "data": [
    {
      "scope": { "type": "user", "user_id": "user_01AbCdEfGh" },
      "amount": "50000",
      "currency": "USD",
      "period": "monthly",
      "source": { "type": "seat_tier", "seat_tier": "enterprise_standard" },
      "spend_limit_id": "spl_01XyZaBcDeFgHiJkLmNoPq",
      "period_to_date_spend": "31402.5"
    }
  ],
  "next_page": "page_..."
}
```

### Validations
- **user_ids[] entry malformed** - 400 - `user_ids[]: entry is not a valid user ID`
- **user_ids[] has more than 100 entries** - 400 - (no message)
- **limit outside 1–1000** - 400 - (no message)
- **page cursor invalid** - 400 - `page: invalid cursor`
- **page cursor does not match current user_ids[]** - 400 - `page: cursor does not match current query parameters`
- **page cursor from a different API version** - 400 - `page: cursor was issued by a different API version`
- **Organization is not on an Enterprise plan** - 400 - `this endpoint is not supported for this organization type`
- **Usage credit billing not enabled** - 400 - `overage billing is not enabled for this organization`
```

--------------------------------

### GET /v1/organizations/spend_limit_increase_requests/{spend_limit_increase_request_id}

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Returns a single spend limit increase request. Requires scope: `read:spend_limits`.

```APIDOC
## GET /v1/organizations/spend_limit_increase_requests/{spend_limit_increase_request_id}

### Description
Returns a single spend limit increase request. Requires scope: `read:spend_limits`.

### Method
GET

### Endpoint
/v1/organizations/spend_limit_increase_requests/{spend_limit_increase_request_id}

### Parameters
#### Path Parameters
- **spend_limit_increase_request_id** (string) - Required - Prefixed `slir_`.

### Request Example
```
curl "https://api.anthropic.com/v1/organizations/spend_limit_increase_requests/slir_01AbCdEfGhIjKlMnOpQrSt" \
  -H "x-api-key: $ANTHROPIC_ADMIN_KEY"
```

### Response
#### Success Response (200)
A SpendLimitIncreaseRequest object.

### Error Handling
#### Validations
- **Condition**: Request not found in this organization - Status: 404 - Message: (none)
- **Condition**: Requester is no longer a member of this organization - Status: 404 - Message: (none)
- **Condition**: Organization is not on an Enterprise plan - Status: 400 - Message: `this endpoint is not supported for this organization type`
- **Condition**: Usage credit billing not enabled - Status: 400 - Message: `overage billing is not enabled for this organization`
```

--------------------------------

### Write tests

Source: https://support.claude.com/en/articles/14553517-claude-code-common-developer-use-cases

Sample prompt for writing tests. Replace [file] and [dir] with the target file and directory.

```text
add tests for @[file] — match the style in [dir]
```

--------------------------------

### GET /v1/organizations/spend_limit_increase_requests

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Retrieves spend limit increase requests for the organization. This endpoint is paginated with an opaque cursor.

```APIDOC
## GET /v1/organizations/spend_limit_increase_requests

### Description
Retrieves spend limit increase requests for the organization. This endpoint is paginated with an opaque cursor.

### Method
GET

### Endpoint
/v1/organizations/spend_limit_increase_requests

### Parameters
#### Query Parameters
- **limit** (integer) - Optional - Maximum number of rows to return in the response.
- **page** (string) - Optional - Opaque cursor for pagination. Pass the `next_page` cursor from the previous response unchanged.
- **user_ids[]** (array of strings) - Optional - Filter results by user IDs. Changing this parameter mid-sequence invalidates the cursor.
- **status[]** (array of strings) - Optional - Filter results by status. Changing this parameter mid-sequence invalidates the cursor.
- **actor_ids[]** (array of strings) - Optional - Filter results by actor IDs. Changing this parameter mid-sequence invalidates the cursor.

### Response
#### Success Response (200)
- **rows** (array) - List of spend limit increase request records.
- **next_page** (string or null) - Opaque cursor for the next page. `null` indicates no more pages.

#### Response Example
{
  "rows": [],
  "next_page": null
}

### Error Handling
- **429 Too Many Requests** - Rate limit exceeded (60 requests per minute per organization).
- **400 Bad Request** - Cursor does not match current query parameters.
```

--------------------------------

### Set model for current session with --model flag

Source: https://support.claude.com/en/articles/11940350

Start Claude Code with the --model flag to set the model for the current session only. Each command uses the exact model identifier for the listed model. The change applies only to that session.

```bash
claude --model claude-sonnet-5-5
```

```bash
claude --model claude-opus-5-5
```

```bash
claude --model claude-fable-5-1
```

```bash
claude --model claude-opus-5
```

```bash
claude --model claude-sonnet-5
```

```bash
claude --model claude-fable-5
```

```bash
claude --model claude-opus-4-8
```

```bash
claude --model claude-opus-4-7
```

```bash
claude --model claude-sonnet-4-6
```

```bash
claude --model claude-opus-4-6
```

```bash
claude --model claude-opus-4-5-20251101
```

```bash
claude --model claude-haiku-4-5-20251001
```

```bash
claude --model claude-sonnet-4-5-20250929
```

--------------------------------

### Prompt: Refactor UserService class in /src/services/user.go

Source: https://support.claude.com/en/articles/12618689-claude-code-on-the-web

Use this prompt to have Claude refactor the UserService class in /src/services/user.go, which is 800 lines long, into three focused services: UserAuthService (login/logout/sessions), UserProfileService (profile CRUD operations), and UserPreferencesService (settings/preferences). Ensure all tests still pass. This approach works well on the web because you can set clear boundaries for Claude to follow.

```text
The UserService class in /src/services/user.go has 800 lines long.  
Split it into three focused services:   
  - UserAuthService (login/logout/sessions)   
  - UserProfileService (profile CRUD operations)  
  - UserPreferencesService (settings/preferences)  
  
Ensure all tests still pass.
```

--------------------------------

### Add M365 MCP Server for Claude service principal via Microsoft Graph

Source: https://support.claude.com/en/articles/12542951-

POST request to add the M365 MCP Server for Claude service principal to your Microsoft Entra tenant. Required for manual setup in Microsoft Entra ID.

```http
POST https://graph.microsoft.com/v1.0/servicePrincipals  
{"appId":"07c030f6-5743-41b7-ba00-0a6e85f37c17"}
```

--------------------------------

### Initialize project memory

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Run /init to generate a CLAUDE.md file at the project root, capturing codebase conventions for future sessions.

```text
/init
```

--------------------------------

### List effective spend limits (GET /v1/organizations/spend_limits/effective)

Source: https://support.claude.com/en/articles/15330651-claude-enterprise-admin-api-reference-guide

Lists every current member of the organization with their resolved effective limit and period-to-date spend. Requires the read:spend_limits scope. Members without a per-user override appear with source.type of seat_tier, rbac_group, or organization; ex-members are not listed.

```http
GET /v1/organizations/spend_limits/effective
```

--------------------------------

### Add M365 MCP Client for Claude service principal via Microsoft Graph

Source: https://support.claude.com/en/articles/12542951-

POST request to add the M365 MCP Client for Claude service principal to your Microsoft Entra tenant. Required for manual setup in Microsoft Entra ID.

```http
POST https://graph.microsoft.com/v1.0/servicePrincipals  
{"appId":"08ad6f98-a4f8-4635-bb8d-f1a3044760f0"}
```

--------------------------------

### Let it handle git

Source: https://support.claude.com/en/articles/14552382-your-first-day-in-claude-code

Ask Claude to stage changes and write a commit message following your repository's conventions.

```text
Stage my changes and write a commit message that follows our existing style.
```

--------------------------------

### Enable Virtual Machine Platform feature

Source: https://support.claude.com/en/articles/12622703-deploy-claude-desktop-for-windows

Run this command to enable the VirtualMachinePlatform feature if it is disabled or if services are missing.

```powershell
Enable-WindowsOptionalFeature -Online -FeatureName VirtualMachinePlatform -All
```

--------------------------------

### Find where something happens

Source: https://support.claude.com/en/articles/14553517-claude-code-common-developer-use-cases

Sample prompt for finding where something happens. Replace [action] with the action you're looking for.

```text
where do we [action] in this codebase?
```
Tokens:
Raw