Gemini ClientCAUTION
A MCP (Model Context Protocol) client that uses Google Gemini AI models for intelligent tool usage and conversation handling. Tested working nicely with Claude Desktop as an MCP Server currently. Based on untested AI gen code by a non-coder use at own risk.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Credit to and started from- medium.com/@fruitful2007/building-an-mcp-client-101-lets-build-one-for-a-gemini-chat-agent
pypi.org/project/gemini-tool-agent
A (potentially, lol) powerful MCP (Model Context Protocol) client that uses Google's Gemini AI models for intelligent tool usage and conversation handling.
- Tested working nicely with Claude Desktop as an MCP Server currently.
Based on untested AI gen code by a non-coder use at own risk.
Features
- Multiple Gemini Models: Support for various Gemini models (2.0-flash, 2.5-pro, 1.5-pro, etc.)
- Flexible Package Support: Works with gemini-tool-agent, google-generativeai, or google-genai packages
- Runtime Model Switching: Change models during conversation
- Server Configuration Management: Store and manage MCP server configurations
- Intelligent Tool Usage: AI-powered tool discovery and execution
- Async/await Support: Efficient concurrent operations
- Comprehensive Error Handling: Robust error recovery and logging
- Type Hints: Modern Python practices with full type safety
Using "adk web" from agents directory- alter the agent.py paths to your own for simple Gemini mcp use for now until I get a better setup. fullmcpagent-README.md
google agent framework report text
Installation
📋 Easy Installation Instructions Here's the correct way to set it up with Claude- maybe, You may need to install this if not working properly:
- cd /gemini-mcp-client && uv add google-generativeai
Doesn't throw an error and works great for chat, not sure about Gemini MCP use. Just edit in y
fa66abe3195bOBSERVED · 2026-10-10Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add mcp-gemini-client --env GEMINI_API_KEY=${GEMINI_API_KEY} --env GOOGLE_API_KEY=${GOOGLE_API_KEY} --env MAX_OUTPUT_TOKENS=${MAX_OUTPUT_TOKENS} -- uvx mcp-gemini-client{
"mcpServers": {
"mcp-gemini-client": {
"command": "uvx",
"args": [
"mcp-gemini-client"
],
"env": {
"GEMINI_API_KEY": "${GEMINI_API_KEY}",
"GOOGLE_API_KEY": "${GOOGLE_API_KEY}",
"MAX_OUTPUT_TOKENS": "${MAX_OUTPUT_TOKENS}"
}
}
}
}Exposed tools (9)
7 read · 2 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
add_numbers | write | Add two numbers together. |
chat_with_gemini | read | |
echo | read | Echo back the provided message. |
echo_message | read | Echo back a message. |
gemini_model_comparison | read | |
get_current_time | read | Get the current time. |
get_server_info | read | Get information about this server. |
list_gemini_models | read | |
set_gemini_model | write |
Trust audit
CAUTIONgrade B · trust 85/100 Install with care. The audit found things worth knowing before you trust its output.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | WARN |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (1 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (6)
reasonflow.db
print(f"API Key: {'Set' if config_manager.client_config.api_key else 'Not set'}")<div id="analysis-${arg.id}" class="analysis" style="display: none;"></div>.pre-commit-config.yaml
.firebaserc
Gates applied: no_behavioural_pass, no_license.
fa66abe3195bfull audit observations/trust-audit/mcp-server/angrysky56__gemini-client.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-10 | fa66abe3195b | CAUTION | B | 85 | first audit |
Questions
What is the Gemini Client MCP server?
A MCP (Model Context Protocol) client that uses Google Gemini AI models for intelligent tool usage and conversation handling. Tested working nicely with Claude Desktop as an MCP Server currently. Based on untested AI gen code by a non-coder use at own risk.
What tools does Gemini Client expose?
9 in total: 7 read-only, 2 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is Gemini Client safe to connect to an agent?
With care. The audit graded it B (85/100) and found 6 things worth knowing before you trust this server, listed below with the exact line each was found on.
What credentials does Gemini Client need?
It reads GEMINI_API_KEY, GOOGLE_API_KEY and MAX_OUTPUT_TOKENS from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Gemini Client run?
It speaks stdio, so it runs as a local process your client starts. It is published on PyPI as mcp-gemini-client.
How current is this page?
The grade is for one exact copy of the source (fa66abe3195b), read on 2026-10-10. The repository is watched and re-audited when it changes.