Convex Migrations logo

Convex Migrations

Community
waynesutton
convex-migrations

Schema migration strategies for evolving applications including adding new fields, backfilling data, removing deprecated fields, index migrations, and zero-downtime migration patterns

Overview

Publisherwaynesutton
Repositoryconvexskills
Skill nameconvex-migrations
Stars
404
Forks
32
Bundled files
3
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by waynesutton on GitHub. Read the source before you install it.

Installation

Install the Convex Migrations AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/waynesutton/convexskills.git /tmp/convexskills
mkdir -p .claude/skills
cp -r /tmp/convexskills/skills/convex-migrations .claude/skills/convex-migrations
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Convex Migrations in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Convex Migrations on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Convex Migrations is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Convex Migrations

Evolve your Convex database schema safely with patterns for adding fields, backfilling data, removing deprecated fields, and maintaining zero-downtime deployments.

Documentation Sources

Before implementing, do not assume; fetch the latest documentation:

Instructions

Migration Philosophy

Convex handles schema evolution differently than traditional databases:

  • No explicit migration files or commands
  • Schema changes deploy instantly with npx convex dev
  • Existing data is not automatically transformed
  • Use optional fields and backfill mutations for safe migrations

Adding New Fields

Start with optional fields, then backfill:

typescript
// Step 1: Add optional field to schema
// convex/schema.ts
import { defineSchema, defineTable } from "convex/server";
import { v } from "convex/values";

export default defineSchema({
  users: defineTable({
    name: v.string(),
    email: v.string(),
    // New field - start as optional
    avatarUrl: v.optional(v.string()),
  }),
});
typescript
// Step 2: Update code to handle both cases
// convex/users.ts
import { query } from "./_generated/server";
import { v } from "convex/values";

export const getUser = query({
  args: { userId: v.id("users") },
  returns: v.union(
    v.object({
      _id: v.id("users"),
      name: v.string(),
      email: v.string(),
      avatarUrl: v.union(v.string(), v.null()),
    }),
    v.null()
  ),
  handler: async (ctx, args) => {
    const user = await ctx.db.get(args.userId);
    if (!user) return null;

    return {
      _id: user._id,
      name: user.name,
      email: user.email,
      // Handle missing field gracefully
      avatarUrl: user.avatarUrl ?? null,
    };
  },
});
typescript
// Step 3: Backfill existing documents
// convex/migrations.ts
import { internalMutation } from "./_generated/server";
import { internal } from "./_generated/api";
import { v } from "convex/values";

const BATCH_SIZE = 100;

export const backfillAvatarUrl = internalMutation({
  args: {
    cursor: v.optional(v.string()),
  },
  returns: v.object({
    processed: v.number(),
    hasMore: v.boolean(),
  }),
  handler: async (ctx, args) => {
    const result = await ctx.db
      .query("users")
      .paginate({ numItems: BATCH_SIZE, cursor: args.cursor ?? null });

    let processed = 0;
    for (const user of result.page) {
      // Only update if field is missing
      if (user.avatarUrl === undefined) {
        await ctx.db.patch(user._id, {
          avatarUrl: generateDefaultAvatar(user.name),
        });
        processed++;
      }
    }

    // Schedule next batch if needed
    if (!result.isDone) {
      await ctx.scheduler.runAfter(0, internal.migrations.backfillAvatarUrl, {
        cursor: result.continueCursor,
      });
    }

    return {
      processed,
      hasMore: !result.isDone,
    };
  },
});

function generateDefaultAvatar(name: string): string {
  return `https://api.dicebear.com/7.x/initials/svg?seed=${encodeURIComponent(name)}`;
}
typescript
// Step 4: After backfill completes, make field required
// convex/schema.ts
export default defineSchema({
  users: defineTable({
    name: v.string(),
    email: v.string(),
    avatarUrl: v.string(), // Now required
  }),
});

Removing Fields

Remove field usage before removing from schema:

typescript
// Step 1: Stop using the field in queries and mutations
// Mark as deprecated in code comments

// Step 2: Remove field from schema (make optional first if needed)
// convex/schema.ts
export default defineSchema({
  posts: defineTable({
    title: v.string(),
    content: v.string(),
    authorId: v.id("users"),
    // legacyField: v.optional(v.string()), // Remove this line
  }),
});

// Step 3: Optionally clean up existing data
// convex/migrations.ts
export const removeDeprecatedField = internalMutation({
  args: {
    cursor: v.optional(v.string()),
  },
  returns: v.null(),
  handler: async (ctx, args) => {
    const result = await ctx.db
      .query("posts")
      .paginate({ numItems: 100, cursor: args.cursor ?? null });

    for (const post of result.page) {
      // Use replace to remove the field entirely
      const { legacyField, ...rest } = post as typeof post & { legacyField?: string };
      if (legacyField !== undefined) {
        await ctx.db.replace(post._id, rest);
      }
    }

    if (!result.isDone) {
      await ctx.scheduler.runAfter(0, internal.migrations.removeDeprecatedField, {
        cursor: result.continueCursor,
      });
    }

    return null;
  },
});

Renaming Fields

Renaming requires copying data to new field, then removing old:

typescript
// Step 1: Add new field as optional
// convex/schema.ts
export default defineSchema({
  users: defineTable({
    userName: v.string(), // Old field
    displayName: v.optional(v.string()), // New field
  }),
});

// Step 2: Update code to read from new field with fallback
export const getUser = query({
  args: { userId: v.id("users") },
  returns: v.object({
    _id: v.id("users"),
    displayName: v.string(),
  }),
  handler: async (ctx, args) => {
    const user = await ctx.db.get(args.userId);
    if (!user) throw new Error("User not found");

    return {
      _id: user._id,
      // Read new field, fall back to old
      displayName: user.displayName ?? user.userName,
    };
  },
});

// Step 3: Backfill to copy data
export const backfillDisplayName = internalMutation({
  args: { cursor: v.optional(v.string()) },
  returns: v.null(),
  handler: async (ctx, args) => {
    const result = await ctx.db
      .query("users")
      .paginate({ numItems: 100, cursor: args.cursor ?? null });

    for (const user of result.page) {
      if (user.displayName === undefined) {
        await ctx.db.patch(user._id, {
          displayName: user.userName,
        });
      }
    }

    if (!result.isDone) {
      await ctx.scheduler.runAfter(0, internal.migrations.backfillDisplayName, {
        cursor: result.continueCursor,
      });
    }

    return null;
  },
});

// Step 4: After backfill, update schema to make new field required
// and remove old field
export default defineSchema({
  users: defineTable({
    // userName removed
    displayName: v.string(),
  }),
});

Adding Indexes

Add indexes before using them in queries:

typescript
// Step 1: Add index to schema
// convex/schema.ts
export default defineSchema({
  posts: defineTable({
    title: v.string(),
    authorId: v.id("users"),
    publishedAt: v.optional(v.number()),
    status: v.string(),
  })
    .index("by_author", ["authorId"])
    // New index
    .index("by_status_and_published", ["status", "publishedAt"]),
});

// Step 2: Deploy schema change
// Run: npx convex dev

// Step 3: Now use the index in queries
export const getPublishedPosts = query({
  args: {},
  returns: v.array(v.object({
    _id: v.id("posts"),
    title: v.string(),
    publishedAt: v.number(),
  })),
  handler: async (ctx) => {
    const posts = await ctx.db
      .query("posts")
      .withIndex("by_status_and_published", (q) =>
        q.eq("status", "published")
      )
      .order("desc")
      .take(10);

    return posts
      .filter((p) => p.publishedAt !== undefined)
      .map((p) => ({
        _id: p._id,
        title: p.title,
        publishedAt: p.publishedAt!,
      }));
  },
});

Changing Field Types

Type changes require careful migration:

typescript
// Example: Change from string to number for a "priority" field

// Step 1: Add new field with new type
// convex/schema.ts
export default defineSchema({
  tasks: defineTable({
    title: v.string(),
    priority: v.string(), // Old: "low", "medium", "high"
    priorityLevel: v.optional(v.number()), // New: 1, 2, 3
  }),
});

// Step 2: Backfill with type conversion
export const migratePriorityToNumber = internalMutation({
  args: { cursor: v.optional(v.string()) },
  returns: v.null(),
  handler: async (ctx, args) => {
    const result = await ctx.db
      .query("tasks")
      .paginate({ numItems: 100, cursor: args.cursor ?? null });

    const priorityMap: Record<string, number> = {
      low: 1,
      medium: 2,
      high: 3,
    };

    for (const task of result.page) {
      if (task.priorityLevel === undefined) {
        await ctx.db.patch(task._id, {
          priorityLevel: priorityMap[task.priority] ?? 1,
        });
      }
    }

    if (!result.isDone) {
      await ctx.scheduler.runAfter(0, internal.migrations.migratePriorityToNumber, {
        cursor: result.continueCursor,
      });
    }

    return null;
  },
});

// Step 3: Update code to use new field
export const getTask = query({
  args: { taskId: v.id("tasks") },
  returns: v.object({
    _id: v.id("tasks"),
    title: v.string(),
    priorityLevel: v.number(),
  }),
  handler: async (ctx, args) => {
    const task = await ctx.db.get(args.taskId);
    if (!task) throw new Error("Task not found");

    const priorityMap: Record<string, number> = {
      low: 1,
      medium: 2,
      high: 3,
    };

    return {
      _id: task._id,
      title: task.title,
      priorityLevel: task.priorityLevel ?? priorityMap[task.priority] ?? 1,
    };
  },
});

// Step 4: After backfill, update schema
export default defineSchema({
  tasks: defineTable({
    title: v.string(),
    // priority field removed
    priorityLevel: v.number(),
  }),
});

Migration Runner Pattern

Create a reusable migration system:

typescript
// convex/schema.ts
import { defineSchema, defineTable } from "convex/server";
import { v } from "convex/values";

export default defineSchema({
  migrations: defineTable({
    name: v.string(),
    startedAt: v.number(),
    completedAt: v.optional(v.number()),
    status: v.union(
      v.literal("running"),
      v.literal("completed"),
      v.literal("failed")
    ),
    error: v.optional(v.string()),
    processed: v.number(),
  }).index("by_name", ["name"]),

  // Your other tables...
});
typescript
// convex/migrations.ts
import { internalMutation, internalQuery } from "./_generated/server";
import { internal } from "./_generated/api";
import { v } from "convex/values";

// Check if migration has run
export const hasMigrationRun = internalQuery({
  args: { name: v.string() },
  returns: v.boolean(),
  handler: async (ctx, args) => {
    const migration = await ctx.db
      .query("migrations")
      .withIndex("by_name", (q) => q.eq("name", args.name))
      .first();
    return migration?.status === "completed";
  },
});

// Start a migration
export const startMigration = internalMutation({
  args: { name: v.string() },
  returns: v.id("migrations"),
  handler: async (ctx, args) => {
    // Check if already exists
    const existing = await ctx.db
      .query("migrations")
      .withIndex("by_name", (q) => q.eq("name", args.name))
      .first();

    if (existing) {
      if (existing.status === "completed") {
        throw new Error(`Migration ${args.name} already completed`);
      }
      if (existing.status === "running") {
        throw new Error(`Migration ${args.name} already running`);
      }
      // Reset failed migration
      await ctx.db.patch(existing._id, {
        status: "running",
        startedAt: Date.now(),
        error: undefined,
        processed: 0,
      });
      return existing._id;
    }

    return await ctx.db.insert("migrations", {
      name: args.name,
      startedAt: Date.now(),
      status: "running",
      processed: 0,
    });
  },
});

// Update migration progress
export const updateMigrationProgress = internalMutation({
  args: {
    migrationId: v.id("migrations"),
    processed: v.number(),
  },
  returns: v.null(),
  handler: async (ctx, args) => {
    const migration = await ctx.db.get(args.migrationId);
    if (!migration) return null;

    await ctx.db.patch(args.migrationId, {
      processed: migration.processed + args.processed,
    });

    return null;
  },
});

// Complete a migration
export const completeMigration = internalMutation({
  args: { migrationId: v.id("migrations") },
  returns: v.null(),
  handler: async (ctx, args) => {
    await ctx.db.patch(args.migrationId, {
      status: "completed",
      completedAt: Date.now(),
    });
    return null;
  },
});

// Fail a migration
export const failMigration = internalMutation({
  args: {
    migrationId: v.id("migrations"),
    error: v.string(),
  },
  returns: v.null(),
  handler: async (ctx, args) => {
    await ctx.db.patch(args.migrationId, {
      status: "failed",
      error: args.error,
    });
    return null;
  },
});
typescript
// convex/migrations/addUserTimestamps.ts
import { internalMutation } from "../_generated/server";
import { internal } from "../_generated/api";
import { v } from "convex/values";

const MIGRATION_NAME = "add_user_timestamps_v1";
const BATCH_SIZE = 100;

export const run = internalMutation({
  args: {
    migrationId: v.optional(v.id("migrations")),
    cursor: v.optional(v.string()),
  },
  returns: v.null(),
  handler: async (ctx, args) => {
    // Initialize migration on first run
    let migrationId = args.migrationId;
    if (!migrationId) {
      const hasRun = await ctx.runQuery(internal.migrations.hasMigrationRun, {
        name: MIGRATION_NAME,
      });
      if (hasRun) {
        console.log(`Migration ${MIGRATION_NAME} already completed`);
        return null;
      }
      migrationId = await ctx.runMutation(internal.migrations.startMigration, {
        name: MIGRATION_NAME,
      });
    }

    try {
      const result = await ctx.db
        .query("users")
        .paginate({ numItems: BATCH_SIZE, cursor: args.cursor ?? null });

      let processed = 0;
      for (const user of result.page) {
        if (user.createdAt === undefined) {
          await ctx.db.patch(user._id, {
            createdAt: user._creationTime,
            updatedAt: user._creationTime,
          });
          processed++;
        }
      }

      // Update progress
      await ctx.runMutation(internal.migrations.updateMigrationProgress, {
        migrationId,
        processed,
      });

      // Continue or complete
      if (!result.isDone) {
        await ctx.scheduler.runAfter(0, internal.migrations.addUserTimestamps.run, {
          migrationId,
          cursor: result.continueCursor,
        });
      } else {
        await ctx.runMutation(internal.migrations.completeMigration, {
          migrationId,
        });
        console.log(`Migration ${MIGRATION_NAME} completed`);
      }
    } catch (error) {
      await ctx.runMutation(internal.migrations.failMigration, {
        migrationId,
        error: String(error),
      });
      throw error;
    }

    return null;
  },
});

Examples

Schema with Migration Support

typescript
// convex/schema.ts
import { defineSchema, defineTable } from "convex/server";
import { v } from "convex/values";

export default defineSchema({
  // Migration tracking
  migrations: defineTable({
    name: v.string(),
    startedAt: v.number(),
    completedAt: v.optional(v.number()),
    status: v.union(
      v.literal("running"),
      v.literal("completed"),
      v.literal("failed")
    ),
    error: v.optional(v.string()),
    processed: v.number(),
  }).index("by_name", ["name"]),

  // Users table with evolved schema
  users: defineTable({
    // Original fields
    name: v.string(),
    email: v.string(),
    
    // Added in migration v1
    createdAt: v.optional(v.number()),
    updatedAt: v.optional(v.number()),
    
    // Added in migration v2
    avatarUrl: v.optional(v.string()),
    
    // Added in migration v3
    settings: v.optional(v.object({
      theme: v.string(),
      notifications: v.boolean(),
    })),
  })
    .index("by_email", ["email"])
    .index("by_createdAt", ["createdAt"]),

  // Posts table with indexes for common queries
  posts: defineTable({
    title: v.string(),
    content: v.string(),
    authorId: v.id("users"),
    status: v.union(
      v.literal("draft"),
      v.literal("published"),
      v.literal("archived")
    ),
    publishedAt: v.optional(v.number()),
    createdAt: v.number(),
    updatedAt: v.number(),
  })
    .index("by_author", ["authorId"])
    .index("by_status", ["status"])
    .index("by_author_and_status", ["authorId", "status"])
    .index("by_publishedAt", ["publishedAt"]),
});

Best Practices

  • Never run npx convex deploy unless explicitly instructed
  • Never run any git commands unless explicitly instructed
  • Always start with optional fields when adding new data
  • Backfill data in batches to avoid timeouts
  • Test migrations on development before production
  • Keep track of completed migrations to avoid re-running
  • Update code to handle both old and new data during transition
  • Remove deprecated fields only after all code stops using them
  • Use pagination for large datasets
  • Add appropriate indexes before running queries on new fields

Common Pitfalls

  1. Making new fields required immediately - Breaks existing documents
  2. Not handling undefined values - Causes runtime errors
  3. Large batch sizes - Causes function timeouts
  4. Forgetting to update indexes - Queries fail or perform poorly
  5. Running migrations without tracking - May run multiple times
  6. Removing fields before code update - Breaks existing functionality
  7. Not testing on development - Production data issues

References

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Convex Migrations AI skill do?

Schema migration strategies for evolving applications including adding new fields, backfilling data, removing deprecated fields, index migrations, and zero-downtime migration patterns

Why use Convex Migrations on TypingMind?

Because you install it once and use it with any model. Convex Migrations is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Convex Migrations in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/waynesutton/convexskills/tree/main/skills/convex-migrations. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Convex Migrations?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Convex Migrations?

As many as you like. As long as a model supports skills, you can use Convex Migrations with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Convex Migrations AI skill free?

Yes. It is published on GitHub by waynesutton under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇