Is Mcp Toolkit safe to install?

SaferSkills independently audited Mcp Toolkit (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.

CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

100

35.0 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 0 flagged

Securityscore 100 · 0 findings

✓—All security checks passedNo findings in this category for the latest scan.pass

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

README.md · 1 file

README.md9.3 KB · Markdown

MCP Toolkit

6 production-ready MCP (Model Context Protocol) servers + CLI discovery tool + Streamlit playground.

MCP is the open standard for connecting AI assistants to external tools. This toolkit provides ready-to-use servers covering common development workflows: database operations, file management, knowledge search, system monitoring, git analytics, and task tracking.

What This Solves

Slow tool integration — Stand up MCP servers in minutes instead of weeks
Fragmented workflows — One toolkit for data, files, git, monitoring, and tasks
Weak governance — Consistent server specs and auto-discovery for safer deployments

Demo

Live demo: https://ct-mcp-toolkit.streamlit.app

Architecture

flowchart TB
    subgraph Client["Client Layer"]
        CLI["CLI (Click)"]
        Playground["Streamlit Playground"]
    end

    subgraph Core["Core"]
        Registry["Server Registry<br/><i>shared/registry.py</i>"]
    end

    subgraph Servers["MCP Servers"]
        FS["file-organizer<br/>5 tools"]
        GH["git-insights<br/>5 tools"]
        KB["markdown-kb<br/>5 tools"]
        AN["system-monitor<br/>5 tools"]
        DB["sqlite-explorer<br/>6 tools"]
        TK["task-tracker<br/>6 tools"]
    end

    subgraph Protocol["Protocol Layer"]
        FastMCP["FastMCP v2<br/><i>stdio / SSE transport</i>"]
    end

    subgraph Execution["Execution Layer"]
        Tools["Tool Execution<br/><i>32 tools across 6 servers</i>"]
    end

    CLI -->|"list / info / serve"| Registry
    Playground -->|"interactive UI"| Registry
    Registry -->|"auto-discovery"| Servers
    Servers -->|"register tools"| FastMCP
    FastMCP -->|"request / response"| Tools

    style Client fill:#e3f2fd,stroke:#1565c0
    style Core fill:#fff3e0,stroke:#ef6c00
    style Servers fill:#e8f5e9,stroke:#2e7d32
    style Protocol fill:#f3e5f5,stroke:#7b1fa2
    style Execution fill:#fce4ec,stroke:#c62828

Directory Layout

mcp-toolkit/
├── mcp_toolkit/
│   ├── servers/           # 6 MCP server implementations
│   │   ├── sqlite_explorer.py    # SQLite CRUD + schema introspection
│   │   ├── file_organizer.py     # File search, dedup, metadata
│   │   ├── markdown_kb.py        # TF-IDF search over markdown docs
│   │   ├── system_monitor.py     # CPU/memory/disk/network via psutil
│   │   ├── git_insights.py       # Repo analytics + contributor stats
│   │   └── task_tracker.py       # Task CRUD with dependency graphs
│   ├── shared/registry.py        # Server auto-discovery
│   └── cli.py                    # CLI: list, info, config, serve
├── app.py                         # Streamlit playground
├── demo_data/                     # Sample files for file-organizer
├── demo_docs/                     # Sample markdown for markdown-kb
└── tests/                         # 180+ pytest tests

Server Catalog

Server	Tools	Description
sqlite-explorer	6	SQLite database CRUD and schema introspection
file-organizer	5	Smart file search, deduplication, metadata extraction
markdown-kb	5	Knowledge base search using TF-IDF over markdown documents
system-monitor	5	Real-time CPU, memory, disk, and network monitoring
git-insights	5	Git repository analytics, blame, contributor statistics
task-tracker	6	Task management with priorities and dependency graphs

Quick Start

Install

git clone https://github.com/ChunkyTortoise/mcp-toolkit.git
cd mcp-toolkit
pip install -e .

Discover Servers

mcp-toolkit list              # List all 6 servers
mcp-toolkit info sqlite-explorer  # Show server tools and details

Configure with Claude Desktop / Claude Code

mcp-toolkit config sqlite-explorer

Output:

{
  "mcpServers": {
    "sqlite-explorer": {
      "command": "mcp-toolkit",
      "args": ["serve", "sqlite-explorer"]
    }
  }
}

Launch a Server

mcp-toolkit serve sqlite-explorer

Try the Playground

make demo

Server Details

sqlite-explorer

Database operations without writing SQL.

Tool	Description
`create_database`	Create DB with tables from schema dict
`query`	Parameterized SELECT queries (rejects mutations)
`insert`	Insert a single row
`update`	Update rows matching WHERE clause
`delete`	Delete rows matching WHERE clause
`get_schema`	Introspect tables, columns, and types

file-organizer

Smart file operations for AI-driven workflows.

Tool	Description
`search_files`	Recursive glob with depth control
`find_duplicates`	SHA-256 content deduplication
`get_metadata`	File size, timestamps, MIME, hash
`analyze_directory`	Size breakdown by extension
`bulk_rename`	Regex rename with dry-run safety

markdown-kb

Search knowledge bases using TF-IDF ranking.

Tool	Description
`index_documents`	Build search index from markdown files
`search`	Cosine similarity ranked search
`get_document`	Retrieve full document content
`list_documents`	List all indexed documents
`get_stats`	Corpus statistics and vocabulary size

system-monitor

Infrastructure observability for AI agents.

Tool	Description
`get_cpu_stats`	Usage percent, cores, frequency
`get_memory_stats`	Total/used/available memory
`get_disk_stats`	Disk usage for any mount point
`get_network_stats`	Network I/O and connection count
`health_check`	Pass/fail against CPU/memory thresholds

git-insights

Repository analytics without complex git commands.

Tool	Description
`get_repo_stats`	Commits, contributors, branches
`get_commit_history`	Recent commits with metadata
`get_blame`	Line-by-line authorship
`get_contributor_stats`	Per-author commit counts
`find_large_files`	Files exceeding size threshold

task-tracker

Task management with dependency resolution.

Tool	Description
`create_task`	Create task with priority (0-4)
`update_task`	Update title, description, status, priority
`delete_task`	Delete task and clean up dependencies
`list_tasks`	Filter by status and priority
`add_dependency`	Add task-to-task dependency
`get_ready_tasks`	Find unblocked pending tasks

Architecture Decisions

ADR	Title	Status
ADR-0001	FastMCP v2 Over Raw Protocol	Accepted
ADR-0002	Six-Server Modularity	Accepted
ADR-0003	Analytics and Cost Tracking	Accepted

Benchmarks

See BENCHMARKS.md for methodology and results. Key numbers:

Operation	P50	Throughput
JSON-RPC Serialize/Deserialize	0.35ms	2,645 ops/sec
Tool Dispatch + Validation	0.06ms	11,987 ops/sec
Server Registry Lookup	0.04ms	15,562 ops/sec
Capability Negotiation	0.96ms	1,000 ops/sec

Development

make test      # Run 180+ pytest tests
make lint      # Ruff lint + format check
make format    # Auto-fix lint + formatting
make demo      # Launch Streamlit playground

Tech Stack

[FastMCP](https://gofastmcp.com) v2 — MCP server framework
Click — CLI interface
Streamlit — Interactive playground
psutil — System monitoring
GitPython — Git operations
scikit-learn — TF-IDF search (markdown-kb)

Key Metrics

Metric	Value
MCP Servers	6
Total Tools	32
Test Suite	180+ pytest tests
Python	3.11+
Protocol	FastMCP v2 (stdio / SSE)
CI	GitHub Actions
License	MIT

Service Mapping

Service 24: Knowledge Management
Service 6: AI-Powered Personal and Business Automation
Service 21: Workflow Optimization

Certification Mapping

IBM RAG and Agentic AI Professional Certificate
Duke University LLMOps Specialization
Google Cloud Generative AI Leader Certificate
Vanderbilt Generative AI Strategic Leader Specialization

Changelog

See CHANGELOG.md for release history.

EnterpriseHub -- Real estate AI platform with BI dashboards and CRM integration
ai-orchestrator -- AgentForge: unified async LLM interface (Claude, Gemini, OpenAI, Perplexity)
llm-integration-starter -- Production LLM patterns: completion, streaming, function calling, RAG, hardening
Portfolio -- Project showcase and services

License

MIT

Mcp Toolkit— agent skill

Is Mcp Toolkit safe to install?

MCP Toolkit

What This Solves

Demo

Architecture

Directory Layout

Server Catalog

Quick Start

Install

Discover Servers

Configure with Claude Desktop / Claude Code

Launch a Server

Try the Playground

Server Details

sqlite-explorer

file-organizer

markdown-kb

system-monitor

git-insights

task-tracker

Architecture Decisions

Benchmarks

Development

Tech Stack

Key Metrics

Service Mapping

Certification Mapping

Changelog

License

Audit the pieces. Scan the whole. Decide.

Is Mcp Toolkit safe to install?

MCP Toolkit

What This Solves

Demo

Architecture

Directory Layout

Server Catalog

Quick Start

Install

Discover Servers

Configure with Claude Desktop / Claude Code

Launch a Server

Try the Playground

Server Details

sqlite-explorer

file-organizer

markdown-kb

system-monitor

git-insights

task-tracker

Architecture Decisions

Benchmarks

Development

Tech Stack

Key Metrics

Service Mapping

Certification Mapping

Changelog

Related Projects

License

Audit the pieces. Scan the whole. Decide.