Devs: LLMs absorb layered complexity until context explodes; tools and MCP are too heavy

dscape · x · 2026-11-21

valigo laments that nobody bothers removing unnecessary complexity layers anymore, since LLMs happily absorb it until the context window explodes. Bradleymeck adds that much implicit knowledge from black-box software design strategies and years of ABI compatibility learning remains undistilled, and tools/MCP are too large to spin up easily within projects.

Original post →

More from coding & agent

coding & agent channel →