Search

> A fork of [mistral-tokenizer-ts](https://github.com/aknorw/mistral-tokenizer) ESM projects compatibility and updated dependencies.

AI Margin Intelligence — one-line, fail-open OTel instrumentation that captures correct, per-attempt LLM cost for the OpenAI / Anthropic Node clients and the Vercel AI SDK.

Pipedream Nyckel Components

Buy-Now-Pay-Later plugin for Better Auth — unified checkout, orders, webhooks, and admin across Tabby and Tamara. Pluggable providers, end-user picker, fully type-safe.

One-command linting suite — ESlint + Prettier + Stylelint + commitlint

tiny text editor with plugins

Fork of langchain-ai/openwiki with an Ollama provider (Ollama Cloud and local).

Privacy-first decentralized storage SDK with pluggable adapters

A tiny, dependency-free, axios-like HTTP client for the browser.

Run anything long-lived under babysit — one supervision substrate for processes and pi subagents, surfaced to the agent as babysit_run/check/send/wait/kill.

Prometheus exporter — scrapes the framework's unified Metrics-API (Effect MetricRegistry) at GET /metrics in Prometheus text exposition format. Counters, histograms (with buckets), gauges + custom metrics; optional bearer-token gate + node process metrics

Cross-agent marketplace search: find, audit, and install plugins/skills/themes/prompts across pi, omp, claude code, opencode, gemini cli, and codex

GateTest — 120-module QA gate with MCP. Eyes (screenshot live pages), ears (production errors), hands (verify fixes worked). Security, supply chain, AI safety, mutation testing. Iterative Claude fix loop. Replaces SonarQube + Snyk + ESLint + 10 others.

Persistence ports for Roborean projects and runs

Annapurna metering hook — emit per-call LLM usage events for per-feature cost attribution.

SQLAlchemy persistence adapter for Roborean

Episteme SDK: A unified framework for advanced uncertainty quantification (KAIROS) and intelligent memory management (MAKS) in LLM applications.

Context apoptosis for LangChain: µs-scale, class-aware history pruning (B.I.O.M.A. kernel) as a drop-in Runnable.

Convert OpenAI Responses/ChatCompletions API request from non-stream mode to stream mode

Convert OpenAI Responses/ChatCompletions API request from non-stream mode to stream mode