agentsclimarketplace

Adobe sdk patterns

Skill jeremylongshore/claude-code-plugins-plus-skills/plugins/saas-packs/adobe-pack/skills/adobe-sdk-patterns

'Apply production-ready patterns for Adobe Firefly Services SDK, PDF Services SDK,From its SKILL.md

Install
npx -y skills add jeremylongshore/claude-code-plugins-plus-skills --skill adobe-sdk-patterns

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its file declares

Copied from the file, not written here

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

7.5 KB, ~1.7k tokens by cl100k_base, as published. Nobody here has run it

Adobe SDK Patterns

Overview

Production-ready patterns for Adobe SDK usage across Firefly Services (@adobe/firefly-apis, @adobe/photoshop-apis, @adobe/lightroom-apis), PDF Services (@adobe/pdfservices-node-sdk), and direct REST API calls.

Prerequisites

  • Completed adobe-install-auth setup
  • Familiarity with async/await patterns
  • Understanding of the Adobe API you are integrating

Instructions

Pattern 1: Singleton Auth Client with Token Caching

// src/adobe/client.ts
import { ServicePrincipalCredentials, PDFServices } from '@adobe/pdfservices-node-sdk';

let pdfServicesInstance: PDFServices | null = null;
let tokenCache: { token: string; expiresAt: number } | null = null;

export function getPDFServices(): PDFServices {
  if (!pdfServicesInstance) {
    const credentials = new ServicePrincipalCredentials({
      clientId: process.env.ADOBE_CLIENT_ID!,
      clientSecret: process.env.ADOBE_CLIENT_SECRET!,
    });
    pdfServicesInstance = new PDFServices({ credentials });
  }
  return pdfServicesInstance;
}

export async function getAccessToken(): Promise<string> {
  if (tokenCache && tokenCache.expiresAt > Date.now() + 300_000) {
    return tokenCache.token;
  }

  const res = await fetch('https://ims-na1.adobelogin.com/ims/token/v3', {
    method: 'POST',
    headers: { 'Content-Type': 'application/x-www-form-urlencoded' },
    body: new URLSearchParams({
      client_id: process.env.ADOBE_CLIENT_ID!,
      client_secret: process.env.ADOBE_CLIENT_SECRET!,
      grant_type: 'client_credentials',
      scope: process.env.ADOBE_SCOPES!,
    }),
  });

  if (!res.ok) throw new Error(`Adobe IMS token error: ${res.status}`);
  const data = await res.json();
  tokenCache = { token: data.access_token, expiresAt: Date.now() + data.expires_in * 1000 };
  return tokenCache.token;
}

Pattern 2: Typed API Wrapper with Error Classification

// src/adobe/firefly-client.ts
export class AdobeApiError extends Error {
  constructor(
    message: string,
    public readonly status: number,
    public readonly code: string,
    public readonly retryable: boolean,
    public readonly retryAfter?: number
  ) {
    super(message);
    this.name = 'AdobeApiError';
  }
}

export async function adobeApiFetch<T>(
  url: string,
  options: RequestInit & { apiKey?: string }
): Promise<T> {
  const token = await getAccessToken();
  const { apiKey, ...fetchOptions } = options;

  const response = await fetch(url, {
    ...fetchOptions,
    headers: {
      'Authorization': `Bearer ${token}`,
      'x-api-key': apiKey || process.env.ADOBE_CLIENT_ID!,
      'Content-Type': 'application/json',
      ...fetchOptions.headers,
    },
  });

  if (!response.ok) {
    const body = await response.text();
    const retryAfter = response.headers.get('Retry-After');
    throw new AdobeApiError(
      `Adobe API ${response.status}: ${body}`,
      response.status,
      response.status === 429 ? 'RATE_LIMITED' :
      response.status === 401 ? 'AUTH_EXPIRED' :
      response.status >= 500 ? 'SERVER_ERROR' : 'CLIENT_ERROR',
      response.status === 429 || response.status >= 500,
      retryAfter ? parseInt(retryAfter) : undefined
    );
  }

  return response.json();
}

Pattern 3: Retry with Exponential Backoff

// src/adobe/retry.ts
export async function withRetry<T>(
  operation: () => Promise<T>,
  config = { maxRetries: 3, baseDelayMs: 1000 }
): Promise<T> {
  for (let attempt = 0; attempt <= config.maxRetries; attempt++) {
    try {
      return await operation();
    } catch (err: any) {
      if (attempt === config.maxRetries) throw err;

      // Only retry on transient errors
      if (err instanceof AdobeApiError && !err.retryable) throw err;

      // Honor Retry-After header from Adobe
      const delay = err.retryAfter
        ? err.retryAfter * 1000
        : config.baseDelayMs * Math.pow(2, attempt) + Math.random() * 500;

      console.warn(`Adobe retry ${attempt + 1}/${config.maxRetries} in ${delay}ms`);
      await new Promise(r => setTimeout(r, delay));
    }
  }
  throw new Error('Unreachable');
}

Pattern 4: Job Polling for Async APIs (Photoshop, Lightroom)

// src/adobe/polling.ts — Photoshop/Lightroom APIs are async (submit job, poll status)
interface AdobeJobStatus {
  status: 'pending' | 'running' | 'succeeded' | 'failed';
  _links?: { self: { href: string } };
  output?: any;
  error?: { code: string; message: string };
}

export async function pollAdobeJob(
  statusUrl: string,
  options = { intervalMs: 2000, timeoutMs: 120_000 }
): Promise<AdobeJobStatus> {
  const token = await getAccessToken();
  const deadline = Date.now() + options.timeoutMs;

  while (Date.now() < deadline) {
    const res = await fetch(statusUrl, {
      headers: {
        'Authorization': `Bearer ${token}`,
        'x-api-key': process.env.ADOBE_CLIENT_ID!,
      },
    });

    const status: AdobeJobStatus = await res.json();

    if (status.status === 'succeeded') return status;
    if (status.status === 'failed') {
      throw new Error(`Adobe job failed: ${status.error?.message || 'Unknown error'}`);
    }

    await new Promise(r => setTimeout(r, options.intervalMs));
  }

  throw new Error('Adobe job polling timeout');
}

Pattern 5: Zod Validation for API Responses

import { z } from 'zod';

const FireflyImageOutputSchema = z.object({
  outputs: z.array(z.object({
    image: z.object({
      url: z.string().url(),
    }),
    seed: z.number(),
  })),
});

const PhotoshopJobSchema = z.object({
  status: z.enum(['pending', 'running', 'succeeded', 'failed']),
  _links: z.object({
    self: z.object({ href: z.string().url() }),
  }).optional(),
});

// Usage
const raw = await adobeApiFetch<unknown>(fireflyUrl, { method: 'POST', body });
const validated = FireflyImageOutputSchema.parse(raw);

Output

  • Type-safe client singleton with token caching
  • Error classification (retryable vs permanent)
  • Automatic retry with Adobe Retry-After header support
  • Async job polling for Photoshop/Lightroom operations
  • Zod runtime validation for API responses

Error Handling

PatternUse CaseBenefit
Token cachingAll API callsAvoids redundant IMS token requests
Error classificationRetry decisionsOnly retries transient failures
Job pollingPhotoshop/LightroomHandles async operation lifecycle
Zod validationAll responsesCatches API contract changes at runtime

Resources

Next Steps

Apply patterns in adobe-core-workflow-a for real-world usage.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,144. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.