Dynamic context discovery

Summary

Cursor describes dynamic context discovery, a coding-agent pattern that provides fewer details initially and lets the agent retrieve relevant information instead of injecting everything as static context. The approach uses files as a simple interface: long tool responses, chat history, Agent Skills, MCP tool descriptions, and integrated terminal output are stored or synchronized so the agent can inspect only what it needs with tools such as tail, grep, rg, and jq. For MCP, only tool names are included initially, while descriptions are synced into one folder per server; an A/B test found a 46.9% reduction in agent tokens for runs that called an MCP tool, with high variance based on the number of MCPs. Cursor says the design reduces context bloat, preserves information truncation could lose, supports recovery after summarization, and can expose MCP re-authentication status. The changes are planned to become available to all users in the coming weeks.

Takeaways

1. 1

   Cursor writes long shell-command and MCP results to files instead of truncating them, allowing agents to inspect the end with tail and read more only when necessary.

2. 2

   Chat history is exposed as a file during summarization, allowing the agent to search for details that were lost from the summary after the context window fills.

3. 3

   In an A/B test of runs that called an MCP tool, dynamically loading MCP descriptions reduced agent tokens by 46.9%; the result had high variance based on the number of MCPs.