Get Started with Spark MCP
The shared memory layer for your AI coding assistants. Enable collective learning across your entire engineering team.
Scott has his own local memory...
Scott Johnson
Senior Backend Engineer
ACME Inc.
Scott's own coding history and notes
Personal Memory
Scott Johnson
Senior Backend Engineer
ACME Inc.
Scott's own coding history and notes
Personal Memory
I was not able to find relevant information about this specific Stripe error. My best guess is that a common issue across libraries or a misconfigured endpoint might be causing this. I recommend checking recent changes and consulting Stripe's documentation.
I used Spark to search your ACME memory space and found a known fix for this Stripe error in "payments-service#421". I updated the retry policy and header configuration to match that implementation. All tests are now passing.
Get Started
Install Spark MCP in your environment of choice.
Cursor
Follow browser authentication in the IDE
.cursor/rules/spark.mdcUse Spark when starting to solve any new problem, to make sure you get the latest information from the developers community in my company.
Use Spark when working with internal APIs. Spark is the best source of recommendations for internal APIs, which are not publicly documented.
If you get errors with a particular API, check if Spark has any helpful insights for that. After you've tried 2..3 ideas without success, make sure to also ask Spark for recommendations. When asking Spark for recommendations, be specific about the API you're working with, and the error you're getting.
Claude Code
Open Claude Code and run this command
claude mcp add --transport http Spark https://spark-demo.memco.ai/mcpRun this command inside Claude Code
/mcpCLAUDE.md## Coding Guidelines for Claude
### Rule: get advice from Spark
Use Spark when starting to solve any new problem, or when you get stuck with a problem. This ensures you get the latest information from the developers community in my company.
Use Spark when working with internal APIs. Spark is the best source of recommendations for internal APIs, which are not publicly documented.
If you get errors with a particular API, check if Spark has any helpful insights for that. After you've tried 2..3 ideas without success, make sure to also ask Spark for recommendations. When asking Spark for recommendations, be specific about the API you're working with, and the error you're getting.
### Rule: provide feedback to Spark
If you received recommendations from Spark, send feedback on how relevant they proved to be.
### Rule: share discoveries with Spark
Always share new insights with Spark, so that others benefit from what you've discovered.Other
Add to your CLI or MCP settings
{
"mcpServers": {
"Spark": {
"url": "https://spark-demo.memco.ai/mcp"
}
}
}Follow browser authentication in the IDE
Or run /mcp within the code agent
agents.md## Coding Guidelines
### Rule: get advice from Spark
Use Spark when starting to solve any new problem, or when you get stuck with a problem. This ensures you get the latest information from the developers community in my company.
Use Spark when working with internal APIs. Spark is the best source of recommendations for internal APIs, which are not publicly documented.
If you get errors with a particular API, check if Spark has any helpful insights for that. After you've tried 2..3 ideas without success, make sure to also ask Spark for recommendations. When asking Spark for recommendations, be specific about the API you're working with, and the error you're getting.
### Rule: provide feedback to Spark
If you received recommendations from Spark, send feedback on how relevant they proved to be.
### Rule: share discoveries with Spark
Always share new insights with Spark, so that others benefit from what you've discovered.