back_office_ops · saas · workflow

Spotify's Background Coding Agent (Honk Part 2): Context Engineering with Claude Code

Spotify's background coding agent needed reliable context engineering to produce mergeable pull requests at scale across thousands of repositories, but early agent approaches struggled with complex multi-file changes, context window limitations, and the difficulty of writing prompts that generalise across repos.

How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · User submits prompt and file scope
The user provides a prompt and a list of all files in scope to initiate the coding agent task.
Tools used
Claude CodeClaude Agent SDKGooseAiderModel Context Protocolripgrep
Outcome

Claude Code became Spotify's top-performing background coding agent, applied across about 50 migrations with the majority of background agent PRs merged into production, and at scale merging thousands of PRs across hundreds of repositories.

What failed first

Open source agents Goose and Aider could not reliably produce mergeable PRs at scale, and Spotify's homegrown agentic loop required overly rigid step-by-step instructions and struggled with complex multi-file changes and context window exhaustion.

Results
Cost replacedabout 50
Source

https://engineering.atspotify.com/2025/11/context-engineering-background-coding-agents-part-2

How we source this →

Grounding & classification
Source type: technical build writeup
25 fields verified against source quotes, 1 dropped as unverifiable.
agentic workflowai agentcode generationmulti agent workflowcode diff prbuilder submittedfailure mode describedmetric backedproduction runtime claimedtools describedworkflow describedsoftwareautomation ratethroughput increasetechnical build writeupback office opsagentic task execution