Summary

This project offers an AI tool generating teaching personas of Buddhist masters from Han, Tibetan, and Southern traditions. It addresses challenges in accessing and understanding Buddhist teachings by providing interactive, citation-backed AI companions, benefiting learners, researchers, and general readers.

How It Works

Leveraging the FoJin platform (503 data sources, 678K+ vectors), the system generates master roles using CBETA, BDRC, and SuttaCentral citations for traceability. A HARD-GATE framework enforces ethical content generation, supported by a two-stage review and automated fidelity testing for accuracy and style.

Quick Start & Requirements

Direct web interaction is available at fojin.app/chat without installation. For developers, installation involves npx master-skill install within AgentSkills-compatible environments (Claude Code, Cursor, etc.), with manual Git clone/pip options. No specific hardware or non-default software is required beyond the AgentSkills environment for CLI usage.

Highlighted Details

• Features 15 pre-defined masters across three traditions, with cross-tradition comparison (/compare-masters).
• Ensures content provenance via mandatory CBETA, BDRC, or SuttaCentral citations.
• Implements a HARD-GATE framework for ethical and content restrictions.
• Offers offline functionality using core text snippets when the FoJin API is unavailable.
• Supports multiple platforms (Claude Code, Cursor, Gemini CLI) and custom master creation (/create-master).

Maintenance & Community

Actively maintained with regular updates (latest v0.6, May 2, 2026). Community interaction via GitHub Discussions; detailed contribution (CONTRIBUTING.md) and ethics (ETHICS.md) guidelines are provided.

Licensing & Compatibility

Released under the MIT License, permitting broad use and compatibility with closed-source applications.

Limitations & Caveats

Generated content is AI-synthesized, not direct historical teachings. Specific esoteric practices are excluded, and the system avoids spiritual attainment judgments. Full functionality relies on the FoJin API, though offline snippets offer a fallback. Adheres strictly to content boundaries, avoiding sectarian promotion or political religious topics.