GGitHubfix(security): resolve Dependabot security vulnerabilities across workspace (#10875)

2fc12056创建于 7月30日历史提交

文件	最后提交记录	最后更新时间
README.md	examples README	1 年前
app.py	Zqjeo upgrade (#2976) * Update the new flutter, ndk version to setup scripts, ci, docs * Reformat app/ * Reformat code for backend, mcp, plugins	11 个月前
dspy_ex.py	Zqjeo upgrade (#2976) * Update the new flutter, ndk version to setup scripts, ci, docs * Reformat app/ * Reformat code for backend, mcp, plugins	11 个月前
langchain_ex.py	examples basics langchain	1 年前
openai_agents_sdk_ex.py	Zqjeo upgrade (#2976) * Update the new flutter, ndk version to setup scripts, ci, docs * Reformat app/ * Reformat code for backend, mcp, plugins	11 个月前
requirements.txt	fix(security): resolve Dependabot security vulnerabilities across workspace (#10875) * fix(backend): update vulnerable python dependencies to patch ranges (~=) * fix(desktop): update vulnerable dependencies in desktop apps to patch ranges (~) * fix(web): update vulnerable dependencies in web applications to patch ranges (~) * fix(sdks): update vulnerable dependencies in react-native sdk and example to patch ranges (~) * fix(mcp): update vulnerable python dependencies in mcp server to patch ranges (~=) * fix(plugins): update vulnerable python dependencies in plugins to patch ranges (~=) * fix(docs): update vulnerable dependencies in docs to patch ranges (~) * fix(omiGlass): update vulnerable dependencies in omiGlass to patch ranges (~) * fix(root): update vulnerable dependencies in root package to patch ranges (~) * fix(windows): apply pnpm security overrides * fix(windows): align electron-builder package family Failure-Class: none Verification: pnpm install --frozen-lockfile; pnpm exec electron-builder install-app-deps; pnpm run typecheck; pnpm run lint; git diff --check. * fix(deps): pin runtime requirements to locked versions Make PR-updated Python runtime requirements exact pins matching their committed backend locks; make MCP's lock-backed constraint explicit and remove its unresolved constraint. Validation: backend/scripts/sync-python-deps.sh; backend/test.sh; mcp uv lock --locked; mcp pytest; locked npm/pnpm resolutions; make preflight. Failure-Class: none	1 个月前

MCP Omi Examples

This repository contains example applications for interacting with the Omi API using MCP (Model Controlled Programs) in different frameworks.

Prerequisites

Before running the examples, make sure you have:

Python 3.8 or later installed
uvx command-line tool installed (pip install uvx)
An OpenAI API key
An Omi UID (Unique Identifier)

Setup

Clone this repository

Create a virtual environment:

python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

Install dependencies:
```
pip install -r requirements.txt
```
Create a .env file in the root directory with the following values:
```
OPENAI_API_KEY=your_openai_api_key
OMI_UID=your_omi_uid
```

Applications

Main Streamlit App

The primary application is a Streamlit chat interface that allows you to interact with your Omi data.

To run the Streamlit app:

streamlit run app.py

The app will open in your browser. You'll need to enter your Omi UID in the sidebar settings before you can start chatting.

Example Scripts

This repository includes three example scripts demonstrating how to interact with Omi data using different frameworks:

DSPy Example (dspy_ex.py):
```
python dspy_ex.py
```
OpenAI Agents SDK Example (openai_agents_sdk_ex.py):
```
python openai_agents_sdk_ex.py
```
LangChain Example (langchain_ex.py):
```
python langchain_ex.py
```

Each example shows a different method of connecting to the Omi API using MCP and can be used as a reference for your own applications.

Features

Access and query your Omi data through natural language
View memories and conversation history
Interactive chat interface (Streamlit app)
Example integrations with popular frameworks (DSPy, OpenAI Agents SDK, LangChain)

Troubleshooting

If you encounter errors about missing uvx, make sure it's installed and in your PATH
Check that both environment variables (OPENAI_API_KEY and OMI_UID) are set correctly
Ensure all dependencies are installed by running pip install -r requirements.txt

OPENAI_API_KEY=
OMI_UID=