Netlify
Frontend deploys, serverless functions.
Verdict
Common use cases
- Check deploy status during incident response
- Trigger production builds from chat
- Review build logs when deploys fail
- Update environment variables across sites
- Verify feature branches deployed before QA
Integration
- Vendor
- Netlify
- Category
- developer-tools
- Auth
- OAUTH2
- Composio slug
netlify
Tools
Setup
Setup guide
- 11. Open your Switchy workspace and navigate to Settings > Integrations > Developer Tools. 2. Click 'Connect' next to Netlify and you'll be redirected to Netlify's OAuth consent screen. 3. Sign in with the Netlify account that owns the sites your team needs to access. 4. Review the requested scopes—typically these include read access to sites and deploys, write access to trigger builds, and manage access for environment variables. 5. Click 'Authorize' to grant Switchy permission. 6. You'll return to Switchy with a confirmation that Netlify is connected. 7. Open any Space, type '@Netlify what's the status of my latest deploy?' and send the message. 8. If the MCP responds with deploy details, the connection works. 9. To invoke specific actions like triggering a build or updating an environment variable, address the MCP directly with '@Netlify' followed by your request.
What teammates see: by default, memories from Netlify are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Check Latest Deploy
@Netlify show me the status of the latest deploy for my main site. Include the deploy time and any error messages if it failed.Open in a Space →
Trigger New Build
@Netlify trigger a new production build for my main site right now. Let me know when the build starts.Open in a Space →
Review Build Logs
@Netlify get the build log for the deploy that failed 20 minutes ago. Show me the last 50 lines so we can see what broke.Open in a Space →
List All Sites
@Netlify list all sites in my account. Include the site name, URL, and last deploy date for each one.Open in a Space →
Update Environment Variable
@Netlify update the environment variable API_KEY to the new value on my production site. Confirm when it's set.Open in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@netlify show me the last 5 deploys for our staging site, including build times and any errors
Here are the last 5 deploys for your staging site: 1. Deploy #847 (main branch) - 2m 14s - Success - deployed 3 hours ago 2. Deploy #846 (feature/auth-fix) - 1m 58s - Success - deployed 5 hours ago 3. Deploy #845 (main branch) - Failed - Build error: missing environment variable API_KEY 4. Deploy #844 (hotfix/cache) - 2m 31s - Success - deployed 1 day ago 5. Deploy #843 (main branch) - 3m 02s - Success - deployed 1 day ago Deploy #845 failed during the build step. The other four completed successfully with build times under 3 minutes.
This example shows deploy history retrieval, a common debugging workflow. Requires OAuth scope for the specific Netlify team and site. The MCP returns structured deploy metadata; the AI formats it for readability. Actual deploy counts and timing depend on your site's activity.
@netlify trigger a new production deploy from the main branch and notify me when it completes
I've triggered a new production deploy from the main branch. Deploy #848 is now building. Current status: Building (step 2 of 4 - installing dependencies) Estimated time remaining: ~2 minutes I'll monitor this deploy and update you when it completes. Note that I can check status but won't receive a push notification — you'll need to ask me to check again in a few minutes, or watch the Netlify dashboard directly.
This example demonstrates triggering deploys programmatically, useful for release workflows. Requires write access via OAuth. The MCP can initiate the deploy but doesn't maintain a persistent connection to watch it — follow-up prompts are needed to poll status. Side-effect: creates a new deploy that consumes build minutes.
@netlify compare the environment variables between our staging and production sites, flag any mismatches that could cause issues
I've compared environment variables across staging and production. Found 3 mismatches: **Missing in production:** - `FEATURE_FLAG_NEW_CHECKOUT` (set to 'true' in staging) - This could cause the new checkout flow to be unavailable in production. **Different values:** - `API_RATE_LIMIT`: staging=1000, production=5000 - Production allows higher throughput. - `LOG_LEVEL`: staging='debug', production='error' - Production suppresses verbose logs. The missing feature flag is the highest risk — if you're testing new checkout in staging, it won't work in production without adding this variable.
This example shows the AI reasoning over data from multiple sites. Requires OAuth access to both environments. The MCP fetches raw environment variable lists; the AI identifies discrepancies and assesses impact. Sensitive variable values may be redacted depending on Netlify's API response format.
Use-case deep-dives
When Netlify MCP closes the loop between dev and design review
A 3-person agency team ships client sites on Netlify and needs designers to QA deploy previews before merging. The MCP lets the dev trigger a deploy from Switchy, grab the preview URL, and drop it in the design thread—all without leaving the conversation. OAuth2 auth means the deploy runs under the dev's Netlify account, so permissions stay clean. This works best when your team already uses Netlify for hosting and you're tired of context-switching to the dashboard just to check build status or fetch a URL. If you're on Vercel or another host, this MCP won't help. The buying call: if you deploy 5+ previews a week and your designers live in Slack or chat, wiring Netlify into Switchy cuts 2 minutes per handoff.
Using Netlify MCP to confirm production deploys in standup
A 6-person product team runs daily standups where the on-call engineer confirms yesterday's production deploy went live and passes basic checks. The Netlify MCP lets them query deploy status, list recent builds, and surface any failed deploys—right in the standup thread. OAuth2 means each engineer authenticates once and the MCP respects their Netlify org access. This scenario wins when you have multiple sites or branches and checking the Netlify UI for each one is tedious. It's less useful if you already pipe deploy webhooks into Slack or if your team is large enough to need a dedicated ops dashboard. The threshold: if you're manually checking Netlify more than twice a day, the MCP pays for itself in saved clicks.
When Netlify MCP helps support roll back a broken deploy fast
A 4-person SaaS support team gets a ticket that the help center is showing a 404. The support lead needs to check if the last deploy broke something and, if so, roll back to the previous build. The Netlify MCP lets them list recent deploys, compare timestamps with the ticket, and trigger a rollback—all from the support chat in Switchy. OAuth2 keeps the rollback action tied to the support lead's Netlify permissions, so there's an audit trail. This works when your support team has Netlify access and you trust them to make deploy decisions without escalating to engineering. If your support team doesn't touch deploys or you have a strict change-control process, the MCP is overkill. The call: if a broken deploy costs you more than 10 minutes of triage, the MCP is worth wiring up.
Frequently asked
What does the Netlify MCP do in Switchy?
It connects your Netlify account so AI agents can read deployment status, site configurations, and build logs without leaving the chat. You can ask questions about failed builds, check which branch is live, or pull environment variables into a conversation. The MCP doesn't deploy sites or change settings — it's read-only visibility into your hosting infrastructure.
Do I need to be a Netlify account owner to connect it?
No, but you need at least Collaborator access to the sites you want the MCP to read. During OAuth setup, Netlify will show which teams and sites you can grant access to. If you're on a free Netlify plan with no team members, you're automatically the owner and can connect immediately.
Can the Netlify MCP trigger new deployments or roll back releases?
No. The MCP is read-only — it surfaces deployment history, build logs, and site metadata but can't start builds, change DNS records, or modify environment variables. If you need to deploy, use Netlify's CLI, dashboard, or Git push triggers. The MCP exists to answer questions, not execute infrastructure changes.
Why use this instead of just opening the Netlify dashboard?
The MCP lets you ask natural-language questions across multiple sites in one chat — "Which of our staging sites failed in the last hour?" or "Show me the build time trend for project X". You don't context-switch to a browser, and the AI can correlate Netlify data with logs from other tools already in the thread.
Does connecting Netlify count against my Switchy plan limits?
MCP connections don't count as seats. You pay for the people using Switchy, not the number of integrations. However, if the AI reads large build logs or pulls data from dozens of sites in one session, that token usage does count toward your workspace's monthly quota.