UI and mcp

[Meetpatel006]

hey anyone know about how seen ui and how can i make connection with mcp , by the way for build i am using the docker composed

[longle325]

Hi, I’m not sure I fully understand your point. Could you clarify what exactly your issues are ? This will help me provide the right support.

[willingWill17]

Problem Disambiguation

To help clarify the setup steps, here are instructions for both Cipher UI and Docker API usage.

Cipher UI Setup

1. Add your OpenAI key to the .env file:

   OPENAI_API_KEY=********

2. Update the memAgent/cipher.yml configuration:

   llm:
     provider: openai
     model: gpt-4o-mini
     apiKey: $OPENAI_API_KEY
     maxIterations: 50
   
   OpenAI:
     embedding:
       type: openai
       model: text-embedding-3-small
       apiKey: $OPENAI_API_KEY

3. Run the following command to start the UI:

   npm install -g @byterover/cipher && cipher --mode ui

Docker API Setup

Please run these commands and provide the logs afterwards:

cd cipher
pnpm i && pnpm run build && npm link
docker compose up -d

After running the commands, please share the result so I can further assist you.

[Meetpatel006]

the docker part is done , but for UI its show me like

[willingWill17]

Windows comes with a built-in command called cipher, which causes a conflict. When you run cipher --mode ui, the system executes the Windows tool instead of the expected npm package.

Setup Instructions

1. Configure your .env file based on the provided .env.example.

2. Start the Cipher UI using npx:

npx @byterover/cipher --mode ui

Let me know if there is still any issue

[Meetpatel006]

for npx installation is working; after installation it throws the same error like New files added to this directory will not be encrypted., but in the code I made some changes by myself, so then it works for the OpenAI API key. When I set other providers like Azure and Gemini, it throws an error like No API key or Ollama configuration found, please set at least one of OPENAI_API_KEY, ANTHROPIC_API_KEY, OPENROUTER_API_KEY, or OLLAMA_BASE_URL in your environment variables \nAvailable providers: OpenAI, Anthropic, OpenRouter, Ollama, Qwen, Gemini, Azure, AWS, LM Studio.

[willingWill17]

I acknowledged the bugs in with using Azure Provider. Since you have Gemini API key, pleas modify the memAgent/cipher.yml file and uncomment the part with gemini models. You config file should look something like this.

llm:
  provider: gemini
  model: gemini-2.5-flash
  apiKey: $GEMINI_API_KEY
  maxIterations: 50
Gemini:
embedding:
  type: gemini
  model: gemini-embedding-001
  apiKey: $GEMINI_API_KEY

If you still face any error, please let me know and i will try to resolve it!