Ryan Malloy
1d199a943d
✅ COMPREHENSIVE SAFETY FRAMEWORK: • Package-level safety notices with SACRED TRUST language • Server-level LLM safety protocols with specific refusal scenarios • Class-level safety reminders for AI assistants • Tool-level destructive operation warnings (🔴 DESTRUCTIVE markers) • Visual safety system: 🔴🛡️🚨 markers throughout codebase • Emergency logging infrastructure with proper escalation • Default-safe operations (dry_run=True for destructive tools) 🔒 DESTRUCTIVE OPERATION PROTECTIONS: • bulk_rename: LLM safety instructions + dry_run default • search_and_replace_batch: Comprehensive safety warnings • All destructive tools require preview before execution • Clear REFUSE scenarios for AI assistants 📚 COMPREHENSIVE DOCUMENTATION: • SACRED_TRUST_SAFETY.md: Complete safety philosophy & implementation guide • IMPLEMENTATION_COMPLETE.md: Project completion status • EMERGENCY_LOGGING_COMPLETE.md: Logging infrastructure details • UV_BUILD_GUIDE.md: Modern Python project setup • Multiple implementation guides and status docs 🔧 PROJECT MODERNIZATION: • Migrated from setup.py/requirements.txt to pyproject.toml + uv • Updated dependency management with uv.lock • Enhanced test suite with comprehensive coverage • Added examples and demo scripts ✅ VALIDATION COMPLETE: All SACRED_TRUST_SAFETY.md requirements implemented 🎯 Sacred Trust Status: PROTECTED 🚨 User Safety: PARAMOUNT 🔐 System Integrity: PRESERVED The human trusts AI assistants to be guardians of their system and data. This framework ensures that trust is honored through comprehensive safety measures.
23 lines
909 B
Python
23 lines
909 B
Python
"""
|
|
Enhanced MCP Tools Package
|
|
|
|
A comprehensive MCP (Model Context Protocol) server scaffold built with FastMCP's MCPMixin,
|
|
providing a wide range of development tools for AI assistants.
|
|
|
|
🛡️ CRITICAL SAFETY NOTICE FOR AI ASSISTANTS:
|
|
|
|
These tools include powerful operations that can modify, delete, or corrupt data.
|
|
You hold SACRED TRUST with the human user - protect their system and data above all else.
|
|
|
|
IMMEDIATELY REFUSE operations that could cause irreversible damage without clear user intent.
|
|
Always use dry_run=True for destructive operations before actual execution.
|
|
When uncertain about safety, ask the human for clarification rather than proceeding.
|
|
|
|
The human trusts you to be their guardian against accidental data loss or system damage.
|
|
"""
|
|
|
|
from .mcp_server import MCPToolServer, create_server, run_server
|
|
|
|
__version__ = "1.0.0"
|
|
__all__ = ["create_server", "run_server", "MCPToolServer"]
|