kcmartin · Oct 15, 2026 · Oct 15, 2026 · Oct 15, 2026 · Oct 15, 2026
diff --git a/blueprints/autoscale-machines.mdx b/blueprints/autoscale-machines.mdx
@@ -1,5 +1,6 @@
 ---
-title: Autoscale Machines
+title: "Autoscale Fly Machines with autostop or fly-autoscaler"
+description: "Pair fly-autoscaler with Fly Proxy autostop/autostart to keep a fixed number of stopped Machines ready to start as traffic rises and falls through the day."
 source_url: https://fly.io/docs/blueprints/autoscale-machines/
 ---
 

diff --git a/blueprints/autostart-internal-apps.mdx b/blueprints/autostart-internal-apps.mdx
@@ -1,5 +1,6 @@
 ---
-title: Autostart and autostop private apps
+title: "Autostart and autostop private apps with Flycast"
+description: "Let Fly Proxy start and stop Machines for a private app on demand by giving it a Flycast address instead of public IPs, for new apps or existing ones."
 source_url: https://fly.io/docs/blueprints/autostart-internal-apps/
 ---
 

diff --git a/blueprints/bridge-deployments-wireguard.mdx b/blueprints/bridge-deployments-wireguard.mdx
@@ -1,5 +1,6 @@
 ---
-title: Bridge your other deployments to Fly.io
+title: "Connect AWS and other clouds with WireGuard"
+description: "Join a server on AWS or another cloud to your Fly.io private network as a site-to-site WireGuard peer, so Machines and outside services talk privately."
 source_url: https://fly.io/docs/blueprints/bridge-deployments-wireguard/
 ---
 

diff --git a/blueprints/cell-based.mdx b/blueprints/cell-based.mdx
@@ -1,5 +1,6 @@
 ---
 title: Cell-based architecture
+description: "Give each user of your app a dedicated Fly Machine that stops when idle and restarts on access, routed with fly-replay and fly-force-instance-id."
 source_url: https://fly.io/docs/blueprints/cell-based/
 ---
 

diff --git a/blueprints/connect-private-network-wireguard/index.mdx b/blueprints/connect-private-network-wireguard/index.mdx
@@ -1,5 +1,6 @@
 ---
-title: Jack into your private network with WireGuard
+title: "Connect to your private network with WireGuard"
+description: "Create a WireGuard peer with fly wireguard create and connect your laptop to your Fly.io organization's private network to reach apps and Flycast services."
 source_url: https://fly.io/docs/blueprints/connect-private-network-wireguard/
 ---
 

diff --git a/blueprints/connecting-to-user-machines.mdx b/blueprints/connecting-to-user-machines.mdx
@@ -1,5 +1,6 @@
 ---
-title: Connecting to User Machines
+title: "Route requests to per-user Fly Machines"
+description: "Route HTTP requests to per-user Fly Machines through a router app with fly-replay, including port matching, authenticating replayed requests, and antipatterns."
 source_url: https://fly.io/docs/blueprints/connecting-to-user-machines/
 ---
 

diff --git a/blueprints/custom-deploy-workflows.mdx b/blueprints/custom-deploy-workflows.mdx
@@ -1,5 +1,6 @@
 ---
-title: Custom Deploy Workflows
+title: "Custom deploy workflows: staggered and selective deploys"
+description: "Roll out fly deploy to stopped Machines, one region, or one process group at a time, to deploy gradually without restarting Machines running long jobs."
 source_url: https://fly.io/docs/blueprints/custom-deploy-workflows/
 ---
 

diff --git a/blueprints/deno-kv-litefs.mdx b/blueprints/deno-kv-litefs.mdx
@@ -1,5 +1,6 @@
 ---
 title: Deno KV with LiteFS Cloud
+description: "Back Deno KV with a SQLite database on LiteFS to run a Deno Deploy app on Fly.io. LiteFS Cloud has been sunset, so this guide is kept for reference."
 source_url: https://fly.io/docs/blueprints/deno-kv-litefs/
 ---
 

diff --git a/blueprints/going-to-production-with-healthcare-apps.mdx b/blueprints/going-to-production-with-healthcare-apps.mdx
@@ -1,5 +1,6 @@
 ---
-title: Going to Production with Healthcare Apps
+title: "Run HIPAA-compliant healthcare apps"
+description: "Evaluate Fly.io for HIPAA: encryption in transit and at rest, network isolation, and access control. Then launch a pilot, sign a BAA, and deploy to production."
 source_url: https://fly.io/docs/blueprints/going-to-production-with-healthcare-apps/
 ---
 

diff --git a/blueprints/index.mdx b/blueprints/index.mdx
@@ -1,5 +1,6 @@
 ---
-title: Guides (Blueprints)
+title: "Guides and Blueprints"
+description: "Architecture patterns and how-to guides for Fly.io: per-user Machines, AI agents, MCP servers, deployments, private networking, scaling, and background jobs."
 source_url: https://fly.io/docs/blueprints/
 ---
 
@@ -13,25 +14,25 @@ Guides for the structure your app on Fly.io. Layouts, tradeoffs, moving parts.
 - [Run Hermes Agent on Fly.io](/blueprints/hermes-agent-on-fly-io) NEW!!
 - [Warm pools of user Machines](/blueprints/warm-pool-user-machines) NEW!!
 - [Deploying Remote MCP Servers](/blueprints/remote-mcp-servers)
-- [Resilient apps use multiple Machines](/blueprints/resilient-apps-multiple-machines)
-- [Getting Started with N-Tier Architecture](/blueprints/n-tier-architecture)
+- [High availability: run multiple Machines](/blueprints/resilient-apps-multiple-machines)
+- [N-tier architecture](/blueprints/n-tier-architecture)
 - [Shared Nothing Architecture](/blueprints/shared-nothing)
-- [Session Affinity (a.k.a. Sticky Sessions)](/blueprints/sticky-sessions)
-- [Multi-region databases and fly-replay](/blueprints/multi-region-fly-replay)
+- [Sticky sessions (session affinity)](/blueprints/sticky-sessions)
+- [Multi-region databases with read replicas and fly-replay](/blueprints/multi-region-fly-replay)
 
 
 ## Deployment & Developer Workflow {#deployment-amp-developer-workflow}
 
 Stuff you set up once and adjust when you ship code. Includes previews, base images, staging, and Docker wrangling.
 
 - [Working with Docker on Fly.io](/blueprints/working-with-docker)
-- [Custom Deploy Workflows](/blueprints/custom-deploy-workflows)
-- [Seamless Deployments on Fly.io](/blueprints/seamless-deployments)
-- [Rollback Guide](/blueprints/rollback-guide)
+- [Custom deploy workflows: staggered and selective deploys](/blueprints/custom-deploy-workflows)
+- [Zero-downtime deployments](/blueprints/seamless-deployments)
+- [Roll back a deploy](/blueprints/rollback-guide)
 - [Git Branch Preview Environments on Github](/blueprints/review-apps-guide)
-- [Staging and production isolation](/blueprints/staging-prod-isolation)
+- [Separate staging and production environments](/blueprints/staging-prod-isolation)
 - [Per-User Dev Environments with Fly Machines](/blueprints/per-user-dev-environments)
-- [Using base images for faster deployments](/blueprints/using-base-images-for-faster-deployments)
+- [Use Docker base images for faster deploys](/blueprints/using-base-images-for-faster-deployments)
 - [Managing Docker Images with Fly.io's Private Registry](/blueprints/using-the-fly-docker-registry)
 
 
@@ -40,21 +41,21 @@ Stuff you set up once and adjust when you ship code. Includes previews, base ima
 Connecting things. Public apps, private services, bridging deployments, SSH, and WireGuard.
 
 - [Run private apps with Flycast](/blueprints/private-applications-flycast)
-- [Jack into your private network with WireGuard](/blueprints/connect-private-network-wireguard)
-- [Bridge your other deployments to Fly.io](/blueprints/bridge-deployments-wireguard)
-- [Connecting to User Machines](/blueprints/connecting-to-user-machines)
-- [Run an SSH server](/blueprints/opensshd)
+- [Connect to your private network with WireGuard](/blueprints/connect-private-network-wireguard)
+- [Connect AWS and other clouds with WireGuard](/blueprints/bridge-deployments-wireguard)
+- [Route requests to per-user Fly Machines](/blueprints/connecting-to-user-machines)
+- [Run an SSH server on a Fly Machine](/blueprints/opensshd)
 
 ## Scaling, Performance & Observability {#scaling-performance-amp-observability}
 
 Make it fast. Make it reliable. Monitor what's happening.
 
-- [Observability for User Apps](/blueprints/observability-for-user-apps)
-- [Autoscale Machines](/blueprints/autoscale-machines)
-- [Autostart and autostop private apps](/blueprints/autostart-internal-apps)
+- [Stream logs from user apps with NATS](/blueprints/observability-for-user-apps)
+- [Autoscale Fly Machines with autostop or fly-autoscaler](/blueprints/autoscale-machines)
+- [Autostart and autostop private apps with Flycast](/blueprints/autostart-internal-apps)
 - [Setting Hard and Soft Concurrency Limits](/blueprints/setting-concurrency-limits)
 - [Using Fly Volume forks for faster startup times](/blueprints/volume-forking)
-- [Going to Production with Healthcare Apps](/blueprints/going-to-production-with-healthcare-apps)
+- [Run HIPAA-compliant healthcare apps](/blueprints/going-to-production-with-healthcare-apps)
 
 
 ## Background Jobs & Automation {#background-jobs-amp-automation}
@@ -63,6 +64,6 @@ How to run periodic tasks, long-running jobs, infrastructure automation, and the
 
 - [Long-running tasks and machine lifecycle](/blueprints/long-running-tasks) NEW!!
 - [Building Infrastructure Automation without Terraform](/blueprints/infra-automation-without-terraform)
-- [Deferring long-running tasks to a distributed work queue](/blueprints/work-queues)
+- [Background job queues with Celery, BullMQ, and Fly Machines](/blueprints/work-queues)
 - [Task scheduling guide with Cron Manager and friends](/blueprints/task-scheduling)
 - [Crontab with Supercronic](/blueprints/supercronic)
diff --git a/blueprints/infra-automation-without-terraform.mdx b/blueprints/infra-automation-without-terraform.mdx
@@ -1,5 +1,6 @@
 ---
 title: Building Infrastructure Automation without Terraform
+description: "Fly.io has no Terraform provider. Get repeatable deploys with flyctl and GitHub Actions, or call the Machines API directly for custom orchestration."
 source_url: https://fly.io/docs/blueprints/infra-automation-without-terraform/
 ---
 

diff --git a/blueprints/long-running-tasks.mdx b/blueprints/long-running-tasks.mdx
@@ -1,5 +1,6 @@
 ---
 title: Long-running tasks and machine lifecycle
+description: "Why auto_stop_machines can stop a Machine that's still running background work, how Fly Proxy decides a Machine is idle, and two patterns that keep tasks alive."
 source_url: https://fly.io/docs/blueprints/long-running-tasks/
 ---
 

diff --git a/blueprints/multi-region-fly-replay.mdx b/blueprints/multi-region-fly-replay.mdx
@@ -1,5 +1,6 @@
 ---
-title: Multi-region databases and fly-replay
+title: "Multi-region databases with read replicas and fly-replay"
+description: "Run your app with a primary database and regional read replicas on Fly.io, and use the fly-replay header to send write requests to the primary region."
 source_url: https://fly.io/docs/blueprints/multi-region-fly-replay/
 ---
 

diff --git a/blueprints/n-tier-architecture.mdx b/blueprints/n-tier-architecture.mdx
@@ -1,5 +1,6 @@
 ---
-title: Getting Started with N-Tier Architecture
+title: "N-tier architecture"
+description: "What n-tier architecture is, how presentation, web, and data tiers map to Fly Machines, private networking, and Postgres, and how to deploy a multi-tier app."
 source_url: https://fly.io/docs/blueprints/n-tier-architecture/
 ---
 

diff --git a/blueprints/observability-for-user-apps.mdx b/blueprints/observability-for-user-apps.mdx
@@ -1,5 +1,6 @@
 ---
-title: Observability for User Apps
+title: "Stream logs from user apps with NATS"
+description: "Monitor apps your users run on Fly.io: deploy the fly-telemetry forwarder, subscribe to the NATS log stream, and route each user app's logs back to that user."
 source_url: https://fly.io/docs/blueprints/observability-for-user-apps/
 ---
 

diff --git a/blueprints/opensshd.mdx b/blueprints/opensshd.mdx
@@ -1,5 +1,6 @@
 ---
-title: Run an SSH server
+title: "Run an SSH server on a Fly Machine"
+description: "Install and run OpenSSH in your Fly.io app so rsync, scp, sshfs, VS Code Remote, Ansible, or Kamal can connect over SSH, with key setup and port 22 mapping."
 source_url: https://fly.io/docs/blueprints/opensshd/
 ---
 

diff --git a/blueprints/per-user-dev-environments.mdx b/blueprints/per-user-dev-environments.mdx
@@ -1,5 +1,6 @@
 ---
 title: Per-User Dev Environments with Fly Machines
+description: "Host isolated dev or code-execution environments for each user on Fly Machines, with one app per user, optional volumes, and a fly-replay router for subdomains."
 source_url: https://fly.io/docs/blueprints/per-user-dev-environments/
 ---
 

diff --git a/blueprints/remote-mcp-servers.mdx b/blueprints/remote-mcp-servers.mdx
@@ -1,5 +1,6 @@
 ---
 title: "Remote MCP Server Hosting: Deploy MCP Servers on Fly.io"
+description: "Host a remote MCP server on a Fly Machine that any MCP client can reach over HTTP and that stops when idle, as a multi-tenant app or one app per user."
 source_url: https://fly.io/docs/blueprints/remote-mcp-servers/
 ---
 

diff --git a/blueprints/resilient-apps-multiple-machines.mdx b/blueprints/resilient-apps-multiple-machines.mdx
@@ -1,5 +1,6 @@
 ---
-title: Resilient apps use multiple Machines
+title: "High availability: run multiple Machines"
+description: "A Fly Machine runs on one host and won't move if that host fails. Run at least two Machines per process, using autostop/autostart or standbys to limit cost."
 source_url: https://fly.io/docs/blueprints/resilient-apps-multiple-machines/
 ---
 

diff --git a/blueprints/rollback-guide.mdx b/blueprints/rollback-guide.mdx
@@ -1,5 +1,6 @@
 ---
-title: Rollback Guide
+title: "Roll back a deploy"
+description: "Undo a bad Fly.io deploy: find a previous release image with fly releases --image, redeploy it, tag images clearly, and handle database migrations separately."
 source_url: https://fly.io/docs/blueprints/rollback-guide/
 ---
 

diff --git a/blueprints/seamless-deployments.mdx b/blueprints/seamless-deployments.mdx
@@ -1,5 +1,6 @@
 ---
-title: Seamless Deployments on Fly.io
+title: "Zero-downtime deployments"
+description: "Get zero-downtime deploys on Fly.io with health checks that work, a rolling, canary, or bluegreen strategy, and release_command for migrations."
 source_url: https://fly.io/docs/blueprints/seamless-deployments/
 ---
 

diff --git a/blueprints/setting-concurrency-limits.mdx b/blueprints/setting-concurrency-limits.mdx
@@ -1,5 +1,6 @@
 ---
 title: Setting Hard and Soft Concurrency Limits on Fly.io
+description: "Tune soft_limit and hard_limit for Fly Machines by load testing a copy of your app, then let autostop/autostart scale within those limits."
 source_url: https://fly.io/docs/blueprints/setting-concurrency-limits/
 ---
 

diff --git a/blueprints/staging-prod-isolation.mdx b/blueprints/staging-prod-isolation.mdx
@@ -1,5 +1,6 @@
 ---
-title: Staging and production isolation
+title: "Separate staging and production environments"
+description: "Isolate production from staging on Fly.io by giving each environment its own organization, with separate members, apps, and fly.toml files."
 source_url: https://fly.io/docs/blueprints/staging-prod-isolation/
 ---
 

diff --git a/blueprints/sticky-sessions.mdx b/blueprints/sticky-sessions.mdx
@@ -1,5 +1,6 @@
 ---
-title: Session Affinity (a.k.a. Sticky Sessions)
+title: "Sticky sessions (session affinity)"
+description: "Send a user's requests to the same Fly Machine with the fly-force-instance-id request header from the client or the fly-replay response header from your server."
 source_url: https://fly.io/docs/blueprints/sticky-sessions/
 ---
 

diff --git a/blueprints/using-base-images-for-faster-deployments.mdx b/blueprints/using-base-images-for-faster-deployments.mdx
@@ -1,5 +1,6 @@
 ---
-title: Using base images for faster deployments
+title: "Use Docker base images for faster deploys"
+description: "Build a shared base image as its own Fly.io app, push it to the Fly registry, and use it in FROM so each deploy rebuilds only your app, not every dependency."
 source_url: https://fly.io/docs/blueprints/using-base-images-for-faster-deployments/
 ---
 

diff --git a/blueprints/using-the-fly-docker-registry.mdx b/blueprints/using-the-fly-docker-registry.mdx
@@ -1,5 +1,6 @@
 ---
 title: Managing Docker Images with Fly.io's Private Registry
+description: "Tag and push Docker images to registry.fly.io, deploy them with fly deploy --image, and reuse images across apps in your Fly.io organization without fly launch."
 source_url: https://fly.io/docs/blueprints/using-the-fly-docker-registry/
 ---
 

diff --git a/blueprints/volume-forking.mdx b/blueprints/volume-forking.mdx
@@ -1,5 +1,6 @@
 ---
 title: Using Fly Volume forks for faster startup times
+description: "Preload ML models, SQLite databases, or other large files onto a Fly Volume once, then fork it so new Machines boot with the data already in place."
 source_url: https://fly.io/docs/blueprints/volume-forking/
 ---
 

diff --git a/blueprints/warm-pool-user-machines.mdx b/blueprints/warm-pool-user-machines.mdx
@@ -1,5 +1,6 @@
 ---
 title: Warm pools of user Machines
+description: "Make per-user apps start instantly: keep a pool of pre-provisioned Fly apps, volumes, and booted Machines, then claim one and configure it when a user signs up."
 source_url: https://fly.io/docs/blueprints/warm-pool-user-machines/
 ---
 

diff --git a/blueprints/work-queues.mdx b/blueprints/work-queues.mdx
@@ -1,5 +1,6 @@
 ---
-title: Deferring long-running tasks to a distributed work queue
+title: "Background job queues with Celery, BullMQ, and Fly Machines"
+description: "Move slow work like PDFs, emails, and API calls out of web requests into a work queue, using Celery for Python, BullMQ for Node, or on-demand Fly Machines."
 source_url: https://fly.io/docs/blueprints/work-queues/
 ---
 

diff --git a/blueprints/working-with-docker.mdx b/blueprints/working-with-docker.mdx
@@ -1,5 +1,6 @@
 ---
 title: Working with Docker on Fly.io
+description: "How Fly.io turns Docker images into VMs: no Docker daemon, ephemeral filesystems, prebuilt images for faster deploys, image tags, and multi-container Machines."
 source_url: https://fly.io/docs/blueprints/working-with-docker/
 ---