diff --git a/.claude/commands/db-check.md b/.claude/commands/db-check.md
new file mode 100644
index 0000000..ba3b05c
--- /dev/null
+++ b/.claude/commands/db-check.md
@@ -0,0 +1,13 @@
+---
+description: Check if database connection is working and show table counts
+---
+
+Check if the database connection is working by:
+
+1. Running `pnpm --filter=@tpmjs/db db:studio` in the background to verify Prisma can connect
+2. If successful, kill the studio process immediately
+3. Report connection status to the user
+
+Then show me the current row counts for all tables (Tool, SyncCheckpoint, SyncLog) by writing a quick script that imports the Prisma client and queries each table.
+
+This helps verify the database is properly configured and shows if any data exists.
diff --git a/.claude/commands/db-migrate.md b/.claude/commands/db-migrate.md
new file mode 100644
index 0000000..d26b76e
--- /dev/null
+++ b/.claude/commands/db-migrate.md
@@ -0,0 +1,12 @@
+---
+description: Create and apply a Prisma migration
+---
+
+Create and apply a new Prisma migration:
+
+1. Run `pnpm --filter=@tpmjs/db db:migrate` to create and apply a migration
+2. The migration will be named automatically based on changes
+3. After migration completes, regenerate the Prisma client with `pnpm --filter=@tpmjs/db db:generate`
+4. Report the results to the user
+
+This is used for production-ready database schema changes that create migration files.
diff --git a/.claude/commands/db-push.md b/.claude/commands/db-push.md
new file mode 100644
index 0000000..329b8a4
--- /dev/null
+++ b/.claude/commands/db-push.md
@@ -0,0 +1,11 @@
+---
+description: Push Prisma schema changes to database (dev only)
+---
+
+Push the current Prisma schema to the database without creating migrations:
+
+1. Run `pnpm --filter=@tpmjs/db db:push` to sync schema changes to the database
+2. After push completes, regenerate the Prisma client with `pnpm --filter=@tpmjs/db db:generate`
+3. Report the results to the user
+
+This is useful for development when you want to quickly iterate on schema changes without creating migration files. Do NOT use this in production - use db:migrate instead.
diff --git a/.claude/commands/db-seed.md b/.claude/commands/db-seed.md
new file mode 100644
index 0000000..eddf444
--- /dev/null
+++ b/.claude/commands/db-seed.md
@@ -0,0 +1,11 @@
+---
+description: Seed the database with initial data
+---
+
+Run the database seed script to initialize the database with default data:
+
+1. Run `pnpm --filter=@tpmjs/db db:seed` to execute the seed script
+2. The seed script will create initial SyncCheckpoint records for the changes feed and keyword sync
+3. Report the results to the user
+
+This should be run once after creating the database to set up the initial sync checkpoints. It's safe to run multiple times - it will only create records if they don't already exist.
diff --git a/.claude/commands/db-studio.md b/.claude/commands/db-studio.md
new file mode 100644
index 0000000..c0d6d94
--- /dev/null
+++ b/.claude/commands/db-studio.md
@@ -0,0 +1,13 @@
+---
+description: Open Prisma Studio to view and edit database data
+---
+
+Open Prisma Studio, a visual database browser:
+
+1. Run `pnpm --filter=@tpmjs/db db:studio` in the background
+2. Wait for the "Prisma Studio is up on http://localhost:5555" message
+3. Tell the user that Prisma Studio is now running at http://localhost:5555
+4. Remind the user they can view and edit all database tables (Tool, SyncCheckpoint, SyncLog) in the browser
+5. Tell them to use Ctrl+C in the terminal or kill the background process when done
+
+This provides a visual interface to browse, search, and edit database records.