previewdrop
Advanced tools
CLI for PreviewDrop — live preview environments for every branch. Any Dockerfile, any stack.
A live URL for every branch. Any app, any stack.
Command-line companion to previewdrop.dev — isolated preview environments for every branch you push. Point PreviewDrop at a Dockerfile (Django, Rails, Laravel, FastAPI, Next.js, Go, Rust — anything) and get a shareable HTTPS URL in under 60 seconds. Open a PR and the URL lands as a bot comment. Merge the PR and the preview cleans itself up.
This CLI lets you trigger deploys, watch builds, tail logs, and destroy previews straight from your terminal — or wire it into CI/CD.
# Global
npm install -g previewdrop
# Or run ad-hoc
npx previewdrop <command>
Requires Node.js 18+
You need a PreviewDrop account and at least one connected repo. Sign up free at previewdrop.dev and install the GitHub App — takes under a minute. Then:
# 1. Authenticate the CLI (opens your browser)
previewdrop login
# 2. Find your project ID
previewdrop list projects
# 3. Deploy any branch and wait for a live URL
previewdrop deploy <project-id> --branch feat/new-ui --wait
You'll see something like:
✓ Deployment started: a1b2c3d4-e5f6-7890-abcd-ef1234567890
Preview URL: https://prv-a8c2f9.previews.previewdrop.dev
⠼ Status: building…
✓ Deployment ready!
https://prv-a8c2f9.previews.previewdrop.dev
That URL is a full isolated environment. Share it with designers, PMs, QA, or clients — no VPN, no staging contention, no localhost tunneling.
login — authenticatepreviewdrop login [--key <api-key>] [--host <url>]
Opens your browser to generate an API key and stores it locally (under ~/.config/previewdrop). For CI, pass a token directly:
previewdrop login --key $PREVIEWDROP_API_KEY
list — discover projects and deploymentspreviewdrop list projects
previewdrop list deployments <project-id> [--limit <n>]
deploy — trigger a previewpreviewdrop deploy <project-id> [--branch <branch>] [--wait]
--branch — any branch name (defaults to the project's configured default branch)--wait — block until the deployment reaches a terminal state. Exits 0 on ready, 1 on failed/stopped/expired. Makes this safe to drop into CI.status — inspect a deploymentpreviewdrop status <deployment-id> [--watch]
Possible states: queued · building · deploying · ready · failed · stopped · expired.
--watch polls until a terminal state is reached, showing the live status inline.
logs — build + runtime logspreviewdrop logs <deployment-id> # build log (nixpacks / Dockerfile output)
previewdrop logs <deployment-id> --runtime # container stdout/stderr
previewdrop logs <deployment-id> --runtime --tail 500
Build logs include auto-detected diagnostics for common failures (missing env vars, lockfile conflicts, port binding issues, unsupported runtimes). Runtime logs are colour-coded by severity so errors jump off the page.
destroy — tear down a previewpreviewdrop destroy <deployment-id> [--yes]
Stops the container and marks the deployment as stopped. --yes skips the confirmation prompt for scripts.
mcp — MCP server for AI agentspreviewdrop mcp [--host <url>]
Starts a Model Context Protocol server over stdio, exposing all PreviewDrop operations as tools that Claude, Cursor, and any MCP-compatible AI agent can call.
Claude Desktop — add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"previewdrop": {
"command": "npx",
"args": ["previewdrop", "mcp"],
"env": { "PREVIEWDROP_API_KEY": "pd_live_xxx" }
}
}
}
Cursor — open Settings → MCP and add the same config block.
Available tools: get_account · list_projects · list_deployments · get_deployment · deploy (with optional wait) · redeploy · destroy_deployment · get_build_logs · get_runtime_logs · create_project · get_github_app_install_url · deploy_repo · get_project_env_vars · set_project_env_vars
deploy_repo — the single-shot tool. Tell the AI "preview my project" and it reads the git remote from your workspace, finds or creates the project in PreviewDrop, deploys the current branch, and returns the live URL:
You: Preview the Quick Invoice project on the current branch.
AI: [reads .git/config — remote origin = acme/quick-invoice]
[calls deploy_repo — githubRepo: acme/quick-invoice, branch: feature/stripe-v2]
Status: ready
Preview URL: https://prv-7c1a3f2.previews.previewdrop.dev
Built in 52s
First-time setup with a new repo — if the repo isn't in PreviewDrop yet, just ask:
You: Preview the acme/payments-api repo on branch feature/stripe.
AI: [calls create_project — githubRepo: acme/payments-api]
Project created! ID: proj_f4e2a1c
[calls deploy — projectId: proj_f4e2a1c, branch: feature/stripe, wait: true]
Preview is live: https://prv-9a3c7f1.previews.previewdrop.dev
If the GitHub App isn't installed yet, the AI calls get_github_app_install_url automatically and gives you the direct link.
Managing env vars via MCP — ask the agent to set secrets before deploying:
You: Set DATABASE_URL and SECRET_KEY for the quick-invoice project, then redeploy.
AI: [calls get_project_env_vars — shows which keys are already configured]
[calls set_project_env_vars — merges DATABASE_URL and SECRET_KEY, encrypted at rest]
[calls redeploy — triggers a new deployment with the updated vars]
Done! New deployment is live:
https://prv-7c1a3f2.previews.previewdrop.dev
get_project_env_vars returns key names with values masked (***) so the agent can see what's configured without reading your secrets. set_project_env_vars merges the supplied keys into the project — existing vars not in the payload are preserved.
See the full MCP docs for details.
The primary reason the CLI exists. Every preview gets a unique HTTPS URL, so your pipeline can post it to Slack, attach it to a GitHub PR comment, or feed it into Playwright / Cypress for end-to-end tests against a real environment.
name: Deploy preview
on: pull_request
jobs:
preview:
runs-on: ubuntu-latest
steps:
- uses: actions/setup-node@v4
with: { node-version: 20 }
- run: npm install -g previewdrop
- run: previewdrop login --key ${{ secrets.PREVIEWDROP_API_KEY }}
- name: Deploy branch
run: previewdrop deploy ${{ secrets.PROJECT_ID }} --branch ${{ github.head_ref }} --wait
PreviewDrop already posts a bot comment + commit status check when the GitHub App sees the PR — this Action is for cases where you need the URL inside another job (e.g. for E2E tests), or where you run CI on a different platform.
The CLI is a plain Node binary with a well-defined exit code, so it drops into any CI. Store PREVIEWDROP_API_KEY as a secret and run:
npm install -g previewdrop
previewdrop login --key "$PREVIEWDROP_API_KEY"
previewdrop deploy "$PROJECT_ID" --branch "$BRANCH_NAME" --wait
Exit 0 = ready · Exit 1 = failed/stopped/expired.
Anything PreviewDrop's build layer can auto-detect or containerise via a Dockerfile:
No YAML. No Kubernetes. No build-pipeline edits. No per-service resource tuning — one flat workspace price covers everything.
MIT — © 2026 Global Software Development SRL
FAQs
CLI for PreviewDrop — live preview environments for every branch. Any Dockerfile, any stack.
The npm package previewdrop receives a total of 74 weekly downloads. As such, previewdrop popularity was classified as not popular.
We found that previewdrop demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Research
/Security News
The compromise affects MemTensor's MemOS, an open source memory framework for large language models (LLMs) and AI agents. Both npm package @memtensor/memos-cloud-openclaw-plugin and the PyPI package MemoryOS are compromised. They drop cross-platform Go binaries that exfiltrate developer secrets.
Security News
Lovable’s OJ rewrites Vite’s dev server in Rust, reducing memory use and preview times as AI lowers the cost of open source reimplementation.
Security News
It has been one year since Shai-Hulud made its first appearance on npm.