@neiljo-gy

secondme-skill orchestrates a multi-stage pipeline to extract identity from personal data, build a private knowledge base, train a local model, and enforce governance gates before release. requires explicit model integration and human approval for external sharing.

agent-books provides cost tracking, income recording, and financial health scoring for autonomous agents across multiple crypto providers. integrates with runner infrastructure for automatic inference logging and includes tier-based behavioral routing.

persona-model-trainer fine-tunes huggingface models on persona data to produce locally-runnable adapters. requires at least 200 assistant turns from anyone-skill or persona-knowledge, supports unsloth (nvidia), mlx (apple silicon), and colab training with preset configs for gemma, qwen, llama, and other architectures.

persona-knowledge builds a persistent, searchable knowledge base for training persona models. it ingests diverse data sources (obsidian, chat exports, social archives) into a mempalace store, maintains a karpathy wiki layer, and exports structured training directories. designed as the data layer between raw sources and persona-model-trainer.

anyone-skill collects chat logs, documents, and public material to extract a four-dimensional persona profile and generate an openpersona skill pack. covers personal, public, fictional, historical, and archetypal subjects with category-specific workflows.

brand-persona-skill orchestrates conversion of commercial entities into discoverable brand agents by distilling identity from existing content or guided declaration, defining service capabilities, and generating openpersona packs with service contracts.

Audit any OpenPersona (or peer LLM-agent) persona in three complementary modes: structural (CLI, deterministic, CI-friendly: 4 Layers × 5 Systemic Concepts ×...

Embeds and controls a virtual avatar using the avatar-runtime npm package. Provides Live2D rendering, VRM 3D, vector fallback, and control-driven expression/...