CLI gives models more control over context than badly designed MCP tools
jobergum · x · 2026-07-25
The author argues that the key difference between a CLI workflow and MCP is control over context management.
- A CLI lets the model use its own context-management skills, such as bash and code mode.
- Poorly designed MCP functions can flood the model with irrelevant context.
- In that case, the model loses the ability to decide what matters and is forced to process noise it cannot control.
More from coding & agent
- HarnessRouter launches with a single API for shipping AI agents in apps — ycombinator · 2026-07-25
- How I built an AI research engine with Claude Opus 5 and web sources — eptwts · 2026-07-25
- OpenHands adds Kimi Code support, metrics, and self-hosted fixes in a community release — rajistics · 2026-07-25
- OpenHands SDK 1.37.1 lets MCP servers add tools while a session is running — rajistics · 2026-07-25
- OpenHands SDK 1.37 adds persistent memory across sessions with MEMORY.md — rajistics · 2026-07-25
- OpenHands brings beta desktop installers and easier skill/plugin recovery — rajistics · 2026-07-25