init

Initialize Sentry in your project (experimental)

Usage

sentry init <target> <directory>

Initialize Sentry in your project (experimental)

Arguments:

Argument	Description
<target>	<org>/<project>, <org>/, <project>, or a directory path
<directory>	Project directory (default: current directory)

Options:

Option	Description
-y, --yes	Non-interactive mode (accept defaults)
-n, --dry-run	Show what would happen without making changes
--features <features>...	Features to enable: errors,tracing,logs,replay,profiling,ai-monitoring,user-feedback
-t, --team <team>	Team slug to create the project under

All commands support --json for machine-readable output and --fields to select specific JSON fields.

Experimental: sentry init is experimental and may modify your source files. Always review changes before committing.

Prerequisites: You must be authenticated first. Run sentry auth login if you haven't already.

Examples

# Interactive setup
sentry init

# Non-interactive with auto-yes
sentry init -y

# Dry run to preview changes
sentry init --dry-run

# Target a subdirectory
sentry init ./my-app

# Use a specific org (auto-detect project)
sentry init acme/

# Use a specific org and project
sentry init acme/my-app

# Assign a team when creating a new project
sentry init acme/ --team backend

# Skip the agent-driven feature picker and use a fixed list (CI / non-interactive)
sentry init --features tracing,replay,sourcemaps

Target Syntax

Syntax	Meaning
(omitted)	Auto-detect org and project
acme/	Use org acme, auto-detect or create project
acme/my-app	Use org acme and project my-app
my-app	Search for project my-app across all accessible orgs

Path-like arguments (starting with ., /, or ~) are always treated as the directory. The order of target and directory can be swapped — the CLI will auto-correct with a warning.

Available Features

Feature	Description
errors	Error monitoring
tracing	Performance tracing
logs	Log integration
replay	Session replay
metrics	Custom metrics
profiling	Profiling
sourcemaps	Source map uploads
crons	Cron job monitoring
ai-monitoring	AI/LLM monitoring
user-feedback	User feedback widget

What the Wizard Does

1. Detects your framework — scans your project files to identify the platform, framework, runtime, and relevant libraries
2. Researches the docs — walks the Sentry docs to figure out which features are supported and useful for your project
3. Asks which features to enable — proposes only the features that fit your stack (e.g. no Session Replay on a server-only Node app), with a short reason next to each one. Error monitoring is always on
4. Installs the SDK — adds the appropriate Sentry SDK package to your project
5. Instruments your code — configures error monitoring and the features you picked

Use --features <list> to skip the analyze-and-pick step (handy for --yes / CI). When provided, the wizard treats it as canonical and the agent goes straight to instrumentation.

Supported Platforms

• JavaScript / TypeScript — Next.js, Express, SvelteKit, React
• Python — Flask, FastAPI

More platforms and frameworks are coming soon.