Back
CursorEngineering PostOfficial SourceagentNew
Agent swarms and the new model economics
Use a strong planner with bounded cheaper workers, separate contexts, shared decisions, and independent review lenses.
CursorJul 20, 20262 min
Source Summary
Cursor’s SQLite experiment suggests swarm harness design can matter more than model mix: separate planner and worker contexts reduced coordination churn while cheaper workers handled most tokens.
Practical Implication
For large agent runs, reserve frontier models for decomposition and decisions, then give bounded implementation work to cheaper workers. Separate their contexts, record shared design decisions, delegate conflict resolution, limit oversized files, and stack independent review lenses.
Agent-Ready Context
Cursor’s new planner-worker swarm reached **73%–85%** of a held-out SQLite test suite within four hours, and every new configuration later reached **100%**. Similar quality across model mixes cost from **$1,339 to $10,565**. For large agent runs, reserve frontier models for decomposition and decisions, then give bounded implementation work to cheaper workers. Separate their contexts, record shared design decisions, delegate conflict resolution, limit oversized files, and stack independent review lenses. This was a controlled but unusual SQLite reimplementation with no source, binary, tests, or internet available to agents. Cursor tested four configurations rather than the full planner-worker matrix, and workers consumed at least **69% of tokens**, so workload shape will strongly affect the economics.
Context Map
agentcoding#multi-agent#harness-engineering#subagentsGeneric AgentPrepare Coding Session
Rate This Item
Personal Note
No note yet. Notes are included in exported bundles.
Related — Every Edge Explained
No approved edges yet for this post.