AI and MCP Guide

The current repo AI story is straightforward: local Ollama or hosted providers through CLI key storage, plus MCP integration that keeps the operator in control.

Provider Setup

zypheron ai start

zypheron ai status

zypheron ai providers

zypheron ai doctor

Hosted Keys Through the CLI

zypheron config set-key anthropic

zypheron config set-key openai

zypheron config set-key deepseek

zypheron config set-key gemini

zypheron config set-key kimi

zypheron config get-providers

The repo’s current guidance is to use the CLI key flow as the preferred user path instead of expecting people to hand-edit environment files for common setups.

Local AI with Ollama

ollama pull llama3.2:3b

ollama pull mistral:latest

zypheron config set ai.provider ollama

zypheron config set ai.model llama3.2:3b

zypheron ai test --provider ollama

MCP Integration

MCP is an integration concept for connecting Zypheron to MCP-capable clients. Verify your local dependencies first, then keep the integration config in the MCP client itself rather than copying stale hardcoded path examples.

zypheron install-deps --all

zypheron config show

TUI model selection

Selecting a hosted model without a configured key should prompt for that key and persist the choice for reuse while keeping the workflow human-in-the-loop.

MCP config hygiene

The safest current pattern is to keep MCP config in the client you control instead of relying on archived hardcoded path examples.

ZYPHERON

ZYPHERON Desktop is a cybersecurity IDE for offensive and defensive workflows. The open source CLI remains available for terminal-first users.

AUTHORIZED USE ONLY

Solutions

Infrastructure

  • CLI Source Code

Network

© 2025 ZYPHERON SYSTEMS//DESKTOP + CLI