Building a Natural Language Interface to the Spotify Ads API with Claude Code Plugins

2026-05-01 · Source: Spotify Engineering · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, medium

Summary

Spotify has released an open-source Claude Code plugin for its Ads API v3, designed to simplify campaign creation through natural language. The plugin translates plain English requests, like "Create an audio campaign called Back to School Promo targeting 25-44 year olds in the US with \$100/day budget," into a sequence of API calls. This system addresses the complexity of the Spotify Ads API, which features over 30 resource types and nested targeting structures. Built using Markdown files, a bash script, and two Python helpers, the plugin leverages Claude Code's architecture where skills, agents, and documentation are all Markdown-based. It avoids Model Context Protocol (MCP) in favor of transparent curl commands and directly uses the 8,600-line OpenAPI v3 spec as the source of truth for API interactions and workflow orchestration via OpenAPI Links.

Key takeaway

For marketing professionals or AI engineers building advertising solutions, this Spotify plugin demonstrates a powerful approach to simplifying complex API interactions. You should consider adopting a natural language interface, grounded by your API's OpenAPI specification, to reduce the cognitive load of multi-step workflows. Prioritize transparent execution, allowing users to inspect and debug API calls, which builds trust and control in automated systems. This method can significantly accelerate campaign deployment and reduce manual orchestration.

Key insights

Natural language interfaces grounded by detailed documentation and transparent execution simplify complex API interactions.

Principles

• Markdown-based plugins enable human-readable, diffable, and version-controllable components.
• OpenAPI specs can serve as the direct source of truth for LLM agents, including workflow graphs via Links.
• Transparent execution via visible curl commands provides auditability and user control.

Method

The agent decomposes natural language requests into sequential API calls, performing value conversions, geo-targeting lookups, pre-flight validations, and orchestrating multi-step entity creation (campaign → ad set → ad) using OpenAPI Links.

In practice

• Create Spotify ad campaigns using conversational natural language prompts.
• Utilize OpenAPI Links to define and automate complex API workflows.
• Build LLM tools with Markdown for easy customization and maintenance.

Topics

• Natural Language Processing
• API Integration
• Claude Code Plugins
• OpenAPI Specification
• Advertising Automation
• LLM Agents

Code references

• spotify/ads-agentic-tools

Best for: NLP Engineer, AI Engineer, Prompt Engineer, Marketing Professional

Related on AIssential

• anthropics / claude-code-action — Claude Code Action automates GitHub PR and issue tasks using Claude AI for code assistance.
• 7 Free Web APIs Every Developer and Vibe Coder Should Know — Integrating web APIs provides AI apps with real-time data for enhanced relevance and reliability.
• Claude Code’s Source Leaks, OpenAI Exits Video Generation, Gemini Adds Music Generation, LLMs Learn at Inference — Voice UIs are poised for widespread adoption, while AI model architectures are evolving for efficiency and new modalities.
• Building a custom Chrome extension using Claude Code — A Chrome extension can automate link summarization from Slack messages using OpenAI.
• v2.1.46 — Claude Code updates focus on performance, new models, and expanded functionality through plugins and agent capabilities.
• Spotify’s new Natural Language API Interface and other Examples Explored — Natural Language Interfaces are expanding beyond user-facing features into internal tools, developer tools, and platform infrastructure.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Spotify Engineering.