osama-fawad/Pekingman

Health v2 · maintenance only

Flags: no_releases young no_license

Adoption not part of the score

1175 stars · 79 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Pekingman is a multimodal AGI-oriented agent system that connects environmental perception, long-term memory, human-like reasoning, emotional consistency, and real-time behavioral response in one continuous agent loop. It targets lifelike virtual characters (NPCs, digital humans, metaverse avatars) with Unity-facing execution and a web monitoring panel.

Use cases

• build lifelike NPCs with persistent long-term memory for simulation games
• create digital humans with consistent emotional expression across long interactions
• run embodied agents that sense and react to changing virtual environments
• prototype AGI-oriented agent loops with perception, memory, planning, and action modules
• connect agent decisions to Unity-based interactive simulations
• monitor agent module activity and shared state via a web panel

When to choose

• you need NPCs or virtual characters with memory continuity and believable behavior over long-running sessions
• you want a modular research stack for embodied, multimodal agent experimentation
• your project uses Unity or simulation platforms and needs an agent decision layer

When to avoid

• you need a production-ready, licensed, well-supported framework (no license is provided)
• you want a lightweight game AI solution for low-complexity worlds where scripted behavior suffices
• you need a mature, battle-tested agent framework with community support

Facets

framework · maturity experimental

agent-framework machine-learning nlp simulation monitoring artificial-intelligence simulation cross-platform game-engine embodied-ai npc-behavior long-term-memory multimodal-perception emotional-consistency unity-integration digital-humans agi-research ai-agents game-development natural-language-processing web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="osama-fawad/Pekingman")