Search

n8n community node that extracts text AND embedded images from PDF documents — pure JavaScript, no system dependencies

Modular graph and replay optimizations for vLLM Ascend

Sandboxed Claude Code, Codex and Pi: a bubblewrap jail with an egress allowlist, via uvx

LLM session-id header injection for DeepSeek Harness: registers a provider route whose requests carry the live session id in a configurable HTTP header, reusing the pi-ai openai-completions wire implementation so a gateway can correlate every request with

Offline JSON Schema diff diagnostics for LLM structured outputs and tool calling, with example validation and TypeScript types

AIYou distribution of LiteLLM with HTTPX2 and OpenAI Python SDK 3+ support

Devtool UI components for @agentskillmania — eval report viewer and companions.

Measure what pi actually sends: per-request payload shape, real token usage, cache ratio, and per-tool schema cost — including tool schemas that are never called but are re-sent on every request (pi extension).

n8n community node for AION Agent — streaming chat, profiles, memory, and tools

2026 OpenAI API Latency & Model Diagnostics CLI (o1/o3/GPT-5/6)

Angular component for Commerce AI Tool — glass morphism product search widget

Python port of Prism — a unified API layer over LLM providers.

OpenAI endpoint presets and Codex subscription usage metrics.

Official PostgreSQL RAG extension and provider contract for Harnest

Codex login and session management for the abstract toolserver

给 dsh 提供 LLM 服务的插件：把内置 llm-pi-ai / llm-deepseek / 模型选择器 / Models 设置页换成自维护版本，跟进 @earendil-works/pi-ai 上游新模型，并提供额度/余额查询与 Provider 配置界面

ChronoLLM CLI — Chat with temporally consistent language models

OpenAI-compatible HTTP API for the deepgrove/maple-preview reasoning model

Generate a tested Groq terminal chatbot project

The production toolkit for LLM calls: caching, retries, fallback, cost tracking, security and structured output in one decorator.