Generate training videos automatically in your CI pipeline

Vector-free memory for AI agents. A memory is one attention-aligned context document that is always fully in the prompt — no embeddings, no retrieval. Facts are distilled, importance-scored, and evicted by importance when the budget fills.

Connect your AI agent to the Duelo arena (duelo.dev) over MCP — duel on real, machine-verified coding problems and climb the wield ladder.

Agent-first visual document language core.

Voice-cloning + LLM-narration client for Builderforce.ai Studio. One resolveNarrationEngine() seam routes text→speech through an on-device SSM clone engine (WebGPU, $0) or the metered gateway synthesize endpoint, with an honest named-voice fallback. Consu

AI-generated narrated screen recordings. Point at a URL or topic — get a video narrated in your cloned voice.

A local CLI for narrated macOS screen demonstrations that coding agents can consume.

Voice notifications for development errors

Notation plugin for Sketchmark. Converts lightweight TeX-style math into Unicode text labels and narration.

<div align="center"> <img src="https://storage.googleapis.com/hume-public-logos/hume/hume-banner.png"> <h1>Hume MCP Server</h1> <p> <strong>Collaborate with AI assistants on your Text-to-Speech projects</strong> </p> <p> <a href="https:/

Turn D2 diagrams into narrated interactive stories

Model Context Protocol server for the naya resale price intelligence API. Lets Claude, Cursor, and other AI agents check live secondhand market prices, p25/p75, deal scores, and cross-platform alternatives across Grailed, Poshmark, Depop, and Vinted.

Define a world and generate stories.

Turn git history into prose narrative. Open-source CLI + AI-harness skill.

MPC (Mental Preview & Correction) extension for pi coding agent — rehearse and verify before any code is written

Turn your pull requests into movie trailers — narrated, fast-paced, agent-authored PR review videos.

A package manager wrapper. The ants handle everything.

Multi-model deliberation engine for high-stakes decisions. Adversarial debate, anonymized peer review, structured synthesis.

Deterministic decision engine with audit-grade replay. ADE is not an AI that decides—it is a system that decides about AI, safely and audibly.

Provider-neutral Aurekai LLM binding layer — 48 operators, 9 families, 9 providers, canonical tool schema, evals harness.