gemini-mcp-tool

An MCP server that exposes tools/commands to let AI assistants call the Gemini CLI (via an MCP client such as Claude Code/Claude Desktop) for natural-language analysis and file/codebase understanding, including a sandbox mode for safer code execution/testing.

Evaluated Mar 30, 2026 (21d ago)
Homepage ↗ Repo ↗ Ai Ml mcp model-context-protocol ai gemini cli code-analysis file-analysis typescript
⚙ Agent Friendliness
53
/ 100
Can an agent use this?
🔒 Security
28
/ 100
Is it safe for agents?
⚡ Reliability
28
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
78
Documentation
60
Error Messages
0
Auth Simplicity
85
Rate Limits
10

🔒 Security

TLS Enforcement
20
Auth Strength
20
Scope Granularity
25
Dep. Hygiene
55
Secret Handling
30

MCP auth, transport, and secret handling are not described in the provided materials; it likely runs locally (so TLS isn’t clearly applicable). Main risks are ungoverned tool execution and inadvertent local file exposure via @ references, plus reliance on Gemini CLI configuration/credentials without documented scoping or data-handling guarantees. Dependency list is relatively standard but no CVE/security posture evidence is provided.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
30
Error Recovery
25
AF Security Reliability

Best When

You want an agent to do local context-aware code/file analysis by delegating to Gemini CLI through MCP, especially for large inputs that benefit from Gemini’s context window.

Avoid When

You cannot guarantee what files/paths the agent may reference or you need strict governance over data egress and tool execution; also avoid if you need well-specified auth/rate-limit/error-code contracts at the MCP layer.

Use Cases

  • Analyze large files or codebases by referencing local paths with @ syntax via the Gemini CLI
  • Ask Gemini natural-language questions about project files
  • Summarize directories or specific files using MCP tool calls
  • Use Gemini CLI sandbox mode to test or run scripts safely (as supported by the underlying Gemini CLI)

Not For

  • As a standalone web/API service for remote clients (it’s intended as a local MCP server integration)
  • Handling sensitive data without understanding Gemini CLI/model data handling and local file exposure risks
  • Production-grade SLAs/uptime guarantees (no evidence of published reliability commitments)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: None documented at MCP layer; relies on underlying Gemini CLI configuration/credentials
OAuth: No Scopes: No

README does not specify any MCP auth mechanism. Authentication/authorization (if any) appears to be handled by the Gemini CLI and its local configuration.

Pricing

Free tier: No
Requires CC: No

Pricing for model usage is not described in the provided README/manifest.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • The tool appears to delegate to Gemini CLI; failures may originate from Gemini CLI invocation, model errors, or sandbox execution behavior rather than clearly surfaced MCP error codes.
  • Agent-driven use of @ file/directory references can accidentally include unintended local files if the agent is not constrained.
  • The README lists slash commands and tools, but does not document the precise MCP schema/tool argument validation, response formats, or error taxonomy.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for gemini-mcp-tool.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered