Skip to content
#

llm-adapter

Here are 23 public repositories matching this topic...

Prajna — on-device memory-augmented reasoning: a 6.7M Cognitive Resonance Network inside frozen Gemma 4 E2B. Passes CEHRI 60/60 via retrieval memory, 91.7% on unseen reworded exams. Trained on a Mac Mini M4, no GPU, no API.

  • Updated Sep 26, 2026
  • Python

模型接入层,只回答一个问题:这套运行调度归谁——工具怎么执行、多步循环怎么走、上下文怎么管。流量由官方 ZCode 二进制签名,不伪造客户端、不提取令牌;工具执行通过 MCP 桥交回 DSH。197 项测试把契约锁死;准入与环境契约两层是 fail-closed,兑现不了的请求在起进程前就拒绝。

  • Updated Oct 26, 2026
  • TypeScript

Add this topic to your repo

To associate your repository with the llm-adapter topic, visit your repo's landing page and select "manage topics."

Learn more