llm-middleware

The toolkit for building motties, models made of code, not weights. LLMs are layer 1. A motty is layer 2, below your agents, and to them it is one more model. Inside, it is your TypeScript. It fits the stack you have, as your gateway or beside it.

Totemheart is a deterministic control kernel for persistent cognition and relational behavior, not a sentiment classifier and not a prompt-engineering trick.

🧠 OpenAI-compatible LLM middleware with chain-of-thought reasoning, ReAct agent loops & self-reflection AI. Drop-in upgrade for any LLM backend.

Flare is an open-source boundary engine for LLMs, enforcing minimal relational safety against synthetic intimacy, identity fusion, and role confusion.

Hermes Agent middleware using TypeSafe Jev for fail-open deterministic fast paths before LLM execution

Per-turn model AND reasoning-effort routing for Hermes Agent on Ollama:Cloud, decided by TypeSafe Jev (a System One decision model, not an LLM). Rewrites the outgoing provider request itself via llm_request middleware. Ollama:Cloud only.

Meanest, Cleanest NLPness! Lightweight LLM middleware for entity risk, sentiment vectors, and Responsibility Futures Analysis.

Portable any-agent Hermes plugin: uncensored render chain (abliteration primary + venice fallback) with dynamic persona cards from agent DNA. Fail-open, any-model, any-profile.