agentleFS
apideck-librariesapideck-libraries/api-skills/skills/apideck-node/SKILL.md

apideck-node

Skill4 starsChanged 3 months agoApache-2.0View on GitHub

Apideck Unified API integration patterns for TypeScript and Node.js. Use when building integrations with accounting software (QuickBooks, Xero, NetSuite), CRMs (Salesforce, HubSpot, Pipedrive), HRIS platforms (Workday, BambooHR), file storage (Google Drive, Dropbox, Box), ATS systems (Greenhouse, Lever), e-commerce, or any of Apideck's 200+ connectors. Covers the @apideck/unify SDK, authentication, CRUD operations, pagination, filtering, webhooks, and Vault connection management.

  • Reads credentials

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 3 are distinctive, and 10 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 10

    Very specific

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

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

    ALWAYS use the @apideck/unify SDK.

    line 37
  2. Very specific:

    ALWAYS pass apiKey, appId, and consumerId when initializing the client.

    line 38
  3. Very specific:

    ALWAYS set the APIDECK_API_KEY environment variable rather than hardcoding API keys.

    line 39
  4. Very specific:

    USE serviceId to specify which downstream connector to use (e.g., "salesforce", "quickbooks", "xero").

    line 40
  5. Very specific:

    USE cursor-based pagination with for await...of for iterating large result sets.

    line 41
  6. Very specific:

    USE the filter parameter to narrow results server-side.

    line 42
  7. Very specific:

    USE the fields parameter to request only the columns you need.

    line 43
  8. Distinctive:

    ALWAYS handle errors with try/catch.

    line 44
  9. Distinctive:

    DO NOT store Apideck API keys, App IDs, or Consumer IDs in source code.

    line 45
  10. Very specific:

    All resources follow the same pattern: apideck.{api}.{resource}.{operation}().

    line 87
  11. Distinctive:

    Use async iteration to automatically handle cursor-based pagination:

    line 133
  12. Very specific:

    When the unified model doesn't cover a connector-specific field, use passThrough:

    line 207
  13. Very specific:

    Use @apideck/vault-js to embed a pre-built modal that lets your users authorize connectors and manage integration settings.

    line 253

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Apideck TypeScript SDK Skill

Overview

The Apideck Unified API provides a single integration layer to connect with 200+ third-party services across accounting, CRM, HRIS, file storage, ATS, e-commerce, and more. The official TypeScript SDK (@apideck/unify) provides typed clients for all unified APIs.

Key capabilities:

  • Accounting - Invoices, bills, payments, ledger accounts, journal entries, tax rates, balance sheets, P&L
  • CRM - Contacts, companies, leads, opportunities, activities, pipelines, notes
  • HRIS - Employees, departments, payrolls, time-off requests, schedules
  • File Storage - Files, folders, drives, shared links, upload sessions
  • ATS - Jobs, applicants, applications
  • Vault - Connection management, OAuth flows, custom field mapping
  • Vault JS - Embeddable modal UI for users to authorize connectors and manage settings
  • Webhook - Event subscriptions and real-time notifications

Installation

npm add @apideck/unify

Requires Node.js 18+. The SDK is fully typed with TypeScript definitions.

IMPORTANT RULES

  • ALWAYS use the @apideck/unify SDK. DO NOT make raw fetch/axios calls to the Apideck API.
  • ALWAYS pass apiKey, appId, and consumerId when initializing the client. These are required for all API calls.
  • ALWAYS set the APIDECK_API_KEY environment variable rather than hardcoding API keys.
  • USE serviceId to specify which downstream connector to use (e.g., "salesforce", "quickbooks", "xero"). If a consumer has multiple connections for an API, serviceId is required.
  • USE cursor-based pagination with for await...of for iterating large result sets. DO NOT implement manual pagination.
  • USE the filter parameter to narrow results server-side. DO NOT fetch all records and filter client-side.
  • USE the fields parameter to request only the columns you need. This reduces response size and improves performance.
  • ALWAYS handle errors with try/catch. The SDK throws typed errors for different HTTP status codes.
  • DO NOT store Apideck API keys, App IDs, or Consumer IDs in source code. Use environment variables or a secrets manager.

Quick Start

import { Apideck } from "@apideck/unify";

const apideck = new Apideck({
  apiKey: process.env["APIDECK_API_KEY"] ?? "",
  appId: "your-app-id",
  consumerId: "your-consumer-id",
});

// List CRM contacts
const { data } = await apideck.crm.contacts.list({
  limit: 20,
  filter: { email: "john@example.com" },
});

for (const contact of data) {
  console.log(contact.name, contact.emails);
}

SDK Patterns

Client Setup

import { Apideck } from "@apideck/unify";

const apideck = new Apideck({
  apiKey: process.env["APIDECK_API_KEY"] ?? "",
  appId: "your-app-id",
  consumerId: "your-consumer-id",
});

The consumerId identifies the end-user whose connections are being used. In multi-tenant apps, set this per-request or per-user session.

CRUD Operations

All resources follow the same pattern: apideck.{api}.{resource}.{operation}().

// LIST - retrieve multiple records
const { data } = await apideck.crm.contacts.list({
  serviceId: "salesforce",
  limit: 20,
  filter: { email: "john@example.com" },
  sort: { by: "updated_at", direction: "desc" },
  fields: "id,name,email,phone_numbers",
});

// CREATE - create a new record
const { data: created } = await apideck.crm.contacts.create({
  serviceId: "salesforce",
  contact: {
    first_name: "John",
    last_name: "Doe",
    emails: [{ email: "john@example.com", type: "primary" }],
    phone_numbers: [{ number: "+1234567890", type: "mobile" }],
  },
});
console.log(created.id); // "contact_abc123"

// GET - retrieve a single record
const { data: contact } = await apideck.crm.contacts.get({
  id: "contact_abc123",
  serviceId: "salesforce",
});

// UPDATE - modify an existing record
const { data: updated } = await apideck.crm.contacts.update({
  id: "contact_abc123",
  serviceId: "salesforce",
  contact: { first_name: "Jane" },
});

// DELETE - remove a record
await apideck.crm.contacts.delete({
  id: "contact_abc123",
  serviceId: "salesforce",
});

Pagination

Use async iteration to automatically handle cursor-based pagination:

const result = await apideck.accounting.invoices.list({
  serviceId: "quickbooks",
  limit: 50,
});

// Automatically fetches next pages
for await (const page of result) {
  for (const invoice of page.data) {
    console.log(invoice.number, invoice.total);
  }
}

Or handle pagination manually:

let cursor: string | undefined;
do {
  const { data, meta } = await apideck.accounting.invoices.list({
    serviceId: "quickbooks",
    limit: 50,
    cursor,
  });
  for (const invoice of data) {
    console.log(invoice.number);
  }
  cursor = meta?.cursors?.next ?? undefined;
} while (cursor);

Error Handling

import { Apideck } from "@apideck/unify";
import * as errors from "@apideck/unify/models/errors";

try {
  const { data } = await apideck.crm.contacts.get({ id: "invalid" });
} catch (e) {
  if (e instanceof errors.BadRequestResponse) {
    console.error("Bad request:", e.message);
  } else if (e instanceof errors.UnauthorizedResponse) {
    console.error("Invalid API key or missing credentials");
  } else if (e instanceof errors.NotFoundResponse) {
    console.error("Record not found");
  } else if (e instanceof errors.PaymentRequiredResponse) {
    console.error("API limit reached or payment required");
  } else if (e instanceof errors.UnprocessableResponse) {
    console.error("Validation error:", e.message);
  } else {
    throw e;
  }
}

Common Parameters

Most list endpoints accept these parameters:

ParameterTypeDescription
serviceIdstringDownstream connector ID (e.g., "quickbooks", "salesforce")
limitnumberMax results per page (1-200, default 20)
cursorstringPagination cursor from previous response
filterobjectResource-specific filter criteria
sortobject{ by: string, direction: "asc" | "desc" }
fieldsstringComma-separated field names to return
passThroughobjectPass-through query parameters for the downstream API

Pass-Through Parameters

When the unified model doesn't cover a connector-specific field, use passThrough:

const { data } = await apideck.accounting.invoices.list({
  serviceId: "quickbooks",
  passThrough: {
    search: "overdue",
  },
});

For creating/updating, use pass_through in the request body to send connector-specific fields:

const { data } = await apideck.crm.contacts.create({
  serviceId: "salesforce",
  contact: {
    first_name: "John",
    last_name: "Doe",
    pass_through: [
      {
        service_id: "salesforce",
        operation_id: "contactsAdd",
        extend_object: { custom_sf_field__c: "value" },
      },
    ],
  },
});

API Namespaces

The SDK organizes APIs by namespace. See the reference files for detailed endpoints:

NamespaceReferenceResources
apideck.accounting.*references/accounting-api.mdinvoices, bills, payments, customers, suppliers, ledgerAccounts, journalEntries, taxRates, creditNotes, purchaseOrders, balanceSheet, profitAndLoss, and more
apideck.crm.*references/crm-api.mdcontacts, companies, leads, opportunities, activities, notes, pipelines, users
apideck.hris.*references/hris-api.mdemployees, companies, departments, payrolls, timeOffRequests
apideck.fileStorage.*references/file-storage-api.mdfiles, folders, drives, driveGroups, sharedLinks, uploadSessions
apideck.ats.*references/ats-api.mdapplicants, applications, jobs
apideck.vault.*references/vault-api.mdconnections, connectionSettings, consumers, customMappings, logs, sessions
apideck.webhook.*references/webhook-api.mdwebhooks, eventLogs

Vault JS (Embeddable UI)

Use @apideck/vault-js to embed a pre-built modal that lets your users authorize connectors and manage integration settings. Session creation must happen server-side.

// 1. Server-side: create a session
const { data } = await apideck.vault.sessions.create({
  session: {
    consumer_metadata: { account_name: "Acme Corp", user_name: "John Doe", email: "john@acme.com" },
    redirect_uri: "https://myapp.com/integrations",
    settings: { unified_apis: ["accounting", "crm"] },
    theme: { vault_name: "My App", primary_color: "#4F46E5" },
  },
});

// 2. Client-side: open the modal
import { ApideckVault } from "@apideck/vault-js";

ApideckVault.open({
  token: sessionToken,
  onConnectionChange: (connection) => console.log("Changed:", connection),
  onClose: () => console.log("Closed"),
});

See references/vault-js.md for full configuration options, theming, React integration, and event callbacks.

Download now (ZIP)

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.