Search

ML task library for TigerFlow

A modular, cross-platform framework for automated DCE-MRI and diffusion MRI research.

Ready-to-use time series models implemented in PyTorch and Lightning.

Snowflake LLM inference library

Portable Claude skill to build beautiful, self-contained HTML presentations: author slides and import brand themes (PowerPoint, image, or description).

An investment research committee that runs inside Claude — analysts, a bull-vs-bear debate, risk audit, trader call, and PM sign-off in one conversation. Research only.

krepis (Greek κρηπίς — the foundation a structure stands on): the MIT foundation library for the Nous Ergon stack — logging, SSM secrets, Telegram/SNS alerts, bounded-backoff HTTP retry, S3 writer locks, the Anthropic payload chokepoint, LLM cost telemetry, trading-calendar/date helpers, and EC2-spot/SSM resilience. Edge-free primitives that every higher layer depends down onto.

Local-first conversation cache, RAG, and routing to reduce cloud LLM spend.

Generic LLM service infrastructure — provider abstraction, retry logic, JSON extraction

Enprompta framework instrumentation for LangChain — emits typed, nested OpenTelemetry spans that Enprompta ingests as first-class Retrieval/Tool/Agent spans.

Enprompta framework instrumentation for the OpenAI SDK — emits typed OpenTelemetry spans that Enprompta ingests as first-class LLM spans.

Local, privacy-first cited-answer + disagreement engine: cites its sources, then shows where they agree and contradict. RAG over a local Ollama model.

A powerful React hook for LocalStorage management with TypeScript support, serialization, and comprehensive event handling

A powerful React hook for SessionStorage management with TypeScript support, serialization, and comprehensive event handling

Boruna: a deterministic, capability-safe programming language and framework for LLM-native applications