agentleFS
decocmsdecocms/studio/.claude/skills/add-mcp-tools/SKILL.md

add-mcp-tools

Skill415 starsChanged 3 months agoMITView on GitHub

Guide for adding new MCP tools with consistent patterns for schemas, tool definitions, registry updates, and Better Auth integration

  • Reads credentials

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 4 are distinctive, and 7 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 7

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    When adding a new domain of tools (e.g., apiKeys, webhooks, secrets), create the following structure:

    line 20
  2. Distinctive:

    Add tool names and metadata:

    line 157
  3. Distinctive:

    If your tools wrap Better Auth APIs, you need to:

    line 256
  4. Distinctive:

    Create test files alongside tool files:

    line 352
  5. Very specific:

    Use the existing test patterns from apps/api/src/tools/connection/ as reference.

    line 362
  6. Very specific:

    Forgetting await ctx.access.check(): - Always check authorization

    line 366
  7. Very specific:

    Missing tool registration: - Update both registry.ts AND tools/index.ts

    line 367
  8. Very specific:

    Inconsistent naming: - Use DOMAIN_ACTION pattern (e.g., API_KEY_CREATE)

    line 368
  9. Very specific:

    Missing default permissions: - Add to auth/index.ts if users should have access by default

    line 369
  10. Distinctive:

    Exposing sensitive data: - Only return secrets at creation time

    line 370
  11. Very specific:

    Missing organization check: - Use requireOrganization(ctx) for org-scoped resources

    line 371

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Adding New MCP Tools to Studio

This guide documents the pattern for adding new MCP tools to Studio codebase. Follow this checklist to ensure consistency with existing tools.

Overview

MCP tools are exposed via the Model Context Protocol and allow programmatic management of resources. Each tool follows a consistent pattern with:

  • Zod schemas for input/output validation
  • defineTool for automatic tracing, metrics, and audit logging
  • Access control via ctx.access.check()
  • Better Auth integration via ctx.boundAuth

File Structure

When adding a new domain of tools (e.g., apiKeys, webhooks, secrets), create the following structure:

apps/api/src/tools/<domain>/
├── schema.ts     # Zod schemas for entities and operations
├── create.ts     # <DOMAIN>_CREATE tool
├── list.ts       # <DOMAIN>_LIST tool
├── update.ts     # <DOMAIN>_UPDATE tool  
├── delete.ts     # <DOMAIN>_DELETE tool
└── index.ts      # Barrel export

Step-by-Step Checklist

1. Create Schema File (schema.ts)

Define Zod schemas for:

  • Entity schema: The resource returned in list/get operations
  • Create input/output schemas: Input for creation, output includes created entity
  • Update input/output schemas: Partial updates
  • Delete input/output schemas: ID input, success confirmation output
// apps/api/src/tools/<domain>/schema.ts
import { z } from "zod";

// Entity schema (what's returned in list operations)
export const MyEntitySchema = z.object({
  id: z.string(),
  name: z.string(),
  // ... other fields
  createdAt: z.union([z.string(), z.date()]),
});

export type MyEntity = z.infer<typeof MyEntitySchema>;

// Create schemas
export const MyCreateInputSchema = z.object({
  name: z.string().min(1).max(255),
  // ... other fields
});

export const MyCreateOutputSchema = z.object({
  item: MyEntitySchema,
});

// List schemas
export const MyListInputSchema = z.object({
  // pagination, filters, etc.
});

export const MyListOutputSchema = z.object({
  items: z.array(MyEntitySchema),
});

// Update schemas
export const MyUpdateInputSchema = z.object({
  id: z.string(),
  // ... partial fields
});

export const MyUpdateOutputSchema = z.object({
  item: MyEntitySchema,
});

// Delete schemas
export const MyDeleteInputSchema = z.object({
  id: z.string(),
});

export const MyDeleteOutputSchema = z.object({
  success: z.boolean(),
  id: z.string(),
});

2. Create Tool Files

Each tool file follows this pattern:

// apps/api/src/tools/<domain>/create.ts
import { defineTool } from "../../core/define-tool";
import { getUserId, requireAuth, requireOrganization } from "../../core/studio-context";
import { MyCreateInputSchema, MyCreateOutputSchema } from "./schema";

export const MY_DOMAIN_CREATE = defineTool({
  name: "MY_DOMAIN_CREATE",
  description: "Create a new resource",

  inputSchema: MyCreateInputSchema,
  outputSchema: MyCreateOutputSchema,

  handler: async (input, ctx) => {
    // 1. Require authentication
    requireAuth(ctx);
    
    // 2. Require organization (if org-scoped)
    const organization = requireOrganization(ctx);
    
    // 3. Check authorization
    await ctx.access.check();
    
    // 4. Get user ID
    const userId = getUserId(ctx);
    if (!userId) {
      throw new Error("User ID required");
    }
    
    // 5. Perform operation
    // Option A: Use ctx.boundAuth for Better Auth operations
    const result = await ctx.boundAuth.myDomain.create({ ... });
    
    // Option B: Use ctx.storage for custom storage operations
    const result = await ctx.storage.myDomain.create({ ... });
    
    // 6. Return result
    return { item: result };
  },
});

3. Create Barrel Export (index.ts)

// apps/api/src/tools/<domain>/index.ts
export { MY_DOMAIN_CREATE } from "./create";
export { MY_DOMAIN_LIST } from "./list";
export { MY_DOMAIN_UPDATE } from "./update";
export { MY_DOMAIN_DELETE } from "./delete";

// Export schemas for external use
export * from "./schema";

4. Update Tool Registry (registry.ts)

Add tool names and metadata:

// apps/api/src/tools/registry.ts

// 1. Add to ToolCategory type (if new category)
export type ToolCategory = "Organizations" | "Connections" | "My Domain";

// 2. Add to ALL_TOOL_NAMES array
const ALL_TOOL_NAMES = [
  // ... existing tools
  "MY_DOMAIN_CREATE",
  "MY_DOMAIN_LIST",
  "MY_DOMAIN_UPDATE",
  "MY_DOMAIN_DELETE",
] as const;

// 3. Add to MANAGEMENT_TOOLS array
export const MANAGEMENT_TOOLS: ToolMetadata[] = [
  // ... existing tools
  {
    name: "MY_DOMAIN_CREATE",
    description: "Create resource",
    category: "My Domain",
  },
  {
    name: "MY_DOMAIN_LIST",
    description: "List resources",
    category: "My Domain",
  },
  {
    name: "MY_DOMAIN_UPDATE",
    description: "Update resource",
    category: "My Domain",
  },
  {
    name: "MY_DOMAIN_DELETE",
    description: "Delete resource",
    category: "My Domain",
    dangerous: true,
  },
];

// 4. Add to TOOL_LABELS
const TOOL_LABELS: Record<ToolName, string> = {
  // ... existing labels
  MY_DOMAIN_CREATE: "Create resource",
  MY_DOMAIN_LIST: "List resources",
  MY_DOMAIN_UPDATE: "Update resource",
  MY_DOMAIN_DELETE: "Delete resource",
};

// 5. Update getToolsByCategory if new category
export function getToolsByCategory() {
  const grouped: Record<string, ToolMetadata[]> = {
    Organizations: [],
    Connections: [],
    "My Domain": [], // Add new category
  };
  // ...
}

5. Register Tools (tools/index.ts)

// apps/api/src/tools/index.ts
import * as MyDomainTools from "./myDomain";

export { MyDomainTools };

export const ALL_TOOLS = [
  // ... existing tools
  MyDomainTools.MY_DOMAIN_CREATE,
  MyDomainTools.MY_DOMAIN_LIST,
  MyDomainTools.MY_DOMAIN_UPDATE,
  MyDomainTools.MY_DOMAIN_DELETE,
] as const satisfies { name: ToolName }[];

6. Add to Default Permissions (if needed)

// apps/api/src/auth/index.ts
apiKey({
  permissions: {
    defaultPermissions: {
      self: [
        // ... existing permissions
        "MY_DOMAIN_LIST", // Read access by default
        // Note: CREATE, UPDATE, DELETE usually NOT default
      ],
    },
  },
}),

7. Add Better Auth Integration (if wrapping Better Auth API)

If your tools wrap Better Auth APIs, you need to:

a. Add types to studio-context.ts:

// apps/api/src/core/studio-context.ts

// Add return types
export type MyDomainCreateResult = Awaited<
  ReturnType<BetterAuthApi["createMyDomain"]>
>;
// ... other types

// Add to BoundAuthClient interface
export interface BoundAuthClient {
  // ... existing methods
  
  myDomain: {
    create(data: { ... }): Promise<MyDomainCreateResult>;
    list(): Promise<MyDomainListResult>;
    update(data: { ... }): Promise<MyDomainUpdateResult>;
    delete(id: string): Promise<void>;
  };
}

b. Implement in context-factory.ts:

// apps/api/src/core/context-factory.ts
function createBoundAuthClient(ctx: AuthContext): BoundAuthClient {
  return {
    // ... existing implementations
    
    myDomain: {
      create: async (data) => {
        return auth.api.createMyDomain({ headers, body: data });
      },
      list: async () => {
        return auth.api.listMyDomain({ headers });
      },
      update: async (data) => {
        return auth.api.updateMyDomain({ headers, body: data });
      },
      delete: async (id) => {
        await auth.api.deleteMyDomain({ headers, body: { id } });
      },
    },
  };
}

Key Patterns to Follow

Authentication & Authorization

// Always in this order:
requireAuth(ctx);                    // 1. Must be authenticated
const org = requireOrganization(ctx); // 2. Must have org context (if org-scoped)
await ctx.access.check();            // 3. Must have permission for this tool

Error Handling

// Throw descriptive errors
if (!result) {
  throw new Error(`Resource not found: ${id}`);
}

if (result.organizationId !== organization.id) {
  throw new Error("Resource not found in organization");
}

Sensitive Data

For sensitive data (like API key values):

// Only return sensitive data at creation time
export const CreateOutputSchema = z.object({
  id: z.string(),
  secretValue: z.string(), // Only here!
});

// Never return in list/get
export const EntitySchema = z.object({
  id: z.string(),
  // NO secretValue field
});

Testing

Create test files alongside tool files:

apps/api/src/tools/<domain>/
├── create.test.ts
├── list.test.ts
├── update.test.ts
└── delete.test.ts

Use the existing test patterns from apps/api/src/tools/connection/ as reference.

Common Mistakes to Avoid

  1. Forgetting await ctx.access.check() - Always check authorization
  2. Missing tool registration - Update both registry.ts AND tools/index.ts
  3. Inconsistent naming - Use DOMAIN_ACTION pattern (e.g., API_KEY_CREATE)
  4. Missing default permissions - Add to auth/index.ts if users should have access by default
  5. Exposing sensitive data - Only return secrets at creation time
  6. Missing organization check - Use requireOrganization(ctx) for org-scoped resources

Download now (ZIP)

  • Other versions2 projects use a version of this.

    Word for word · 2

    1. decocms/admin★ 409100% same
    2. decocms/mesh★ 409100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository4 other files it gives its agents.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.