SYSTEM Cited by 1 source

Pinterest Spark MCP Server

Definition

Pinterest Spark MCP Server is one of the three named seed servers in Pinterest's MCP ecosystem, exposing Apache Spark debugging capabilities as agent-callable tools (Source: sources/2026-03-19-pinterest-building-an-mcp-ecosystem-at-pinterest).

Role

"Spark MCP server: underpins our AI Spark debugging experience, used to diagnose Spark job failures, summarize logs, and help record structured root-cause analyses, turning noisy operational threads into reusable knowledge."

Three named functions:

1. Diagnose Spark job failures — agents pull failure signals (driver logs, executor logs, stage DAG) via tool calls.
2. Summarise logs — LLM-backed summarisation of noisy operational log output.
3. Record structured root-cause analyses — the RCA artefact itself is stored via the server, building an on-wiki knowledge base from each incident.

Channel-scoped tool visibility

Pinterest's AI bots on the internal chat platform restrict certain MCP tools to specific communication channels:

"it also supports functionality such as restricting certain MCP tools to certain communication channels (for example, Spark MCP tools are only available in Airflow support channels)."

Canonical wiki instance of per-channel tool visibility as an agent-guardrail mechanism — the Spark MCP tools are visible to agents only in the Airflow support channel, reducing accidental use in unrelated conversations.

Seen in

• sources/2026-03-19-pinterest-building-an-mcp-ecosystem-at-pinterest — canonical wiki instance. Named as one of three seed high-leverage MCP servers, with knowledge-harvesting framing ("turning noisy operational threads into reusable knowledge").

Related

• systems/model-context-protocol — the protocol.
• systems/pinterest-mcp-registry — the registry this server lists in.
• patterns/hosted-mcp-ecosystem — umbrella pattern.
• patterns/layered-jwt-plus-mesh-auth — the auth layering.
• companies/pinterest.