Show HN: Jev in action – task classifier and auto model router in Keen Code

Category: ai-ml

Tags: task-classification, model-routing, decision-models, coding-assistant, llm

Score: 5.7/10 (Innovation: 6, Technical: 6, Documentation: 7, Utility: 5)

Keen Code implements a task complexity classifier using typed decision models, which enables automatic routing between simple, standard, and complex chat models based on user turns. It features bounded context serialization, fail-open behavior, and a clear separation between classification signals and routing policy, making it an interesting approach to adaptive model selection in coding assistants.

Target audience: AI engineers and developers building LLM-powered coding assistants

Repository: https://mochow13.github.io/keen-code/docs/task-classification-using-decision-models.html

View on Hacker News