Search

MiMo-V2.5 is a native omnimodal model by Xiaomi. It delivers Pro-level agentic performance at roughly half the inference cost, while surpassing MiMo-V2-Omni in multimodal perception across image and video understanding...

Trinity-Large-Preview is a frontier-scale open-weight language model from Arcee, built as a 400B-parameter sparse Mixture-of-Experts with 13B active parameters per token using 4-of-256 expert routing. It excels in creative writing,...

Python SDK for Coproduct A2A — provable-coordination extension client with bearer-token minting, manifest signature verification, and framework adapters (LangGraph, CrewAI).

Static linter for LLM prompt scaffolds — catches oversized scaffolds and incompatible technique mixes (evidential + step-by-step + format).

SandrPod sandbox integration for Deep Agents

A hierarchical multi-agent Deep Agent harness built on Pydantic AI

PayClaw tools for LangChain and LangGraph — give your agent a USDC wallet on Base it can actually spend. Gasless via Circle Paymaster.

PayClaw tools for CrewAI — give your agent crew a USDC wallet on Base it can actually spend. Gasless via Circle Paymaster.

Agent framework for remediating GitHub security alerts.

6-layer intervention framework for cognitive bases — structural enforcement over procedural rules

AIOPE — 46-tool AI agent for Android. Linux terminal, browser automation, SSH, dynamic UI, MCP, any model. Built by one developer + AI in 30 days.

Production-ready n8n workflow template: a natural-language data analyst chatbot over Supabase/Postgres. Built with LangChain Tools Agent + Claude Sonnet 4.6 + QuickChart.

PayClaw tools for Microsoft AutoGen — give your AssistantAgent a USDC wallet on Base it can actually spend. Gasless via Circle Paymaster.

Three-model planning debate framework: Claude × Codex × OpenRouter independently plan, cross-review, revise, and one judges — catching confident-but-wrong architectural decisions before they ship.

A governed AI agent runtime where verification comes before trust—typed tasks, approval gates, receipt chains, and controlled forge execution with inspectable behavior.

An advanced Model Context Protocol (MCP) server that turns LinkedIn into an API for LLMs and AI Agents. Features stealth browser automation, headless auth, zero-config component registry, and 24+ tools for jobs, CRM, and profile extraction.

CrewAI integration for Kerq trust verification — score AI tools before your agents connect.

Portable, self-hosting Spec-Driven Development kit for LLM-agent-driven repos. Works across Claude Code, Codex, Cursor, Copilot, Windsurf, and more.

Simple long-running Claude agent CLI

Simple long-running Claude agent CLI