Atlas / MCP servers / zachhandley / ZMCPTools

ZMCPToolsBLOCK

Name: ZMCPTools
Author: zachhandley

mcp/zachhandley/zmcptools

MCP server by zachhandley · claude · claude ai · claude code · Audited 2026-10-08

⚑ Report / re-scan Source ↗

A custom TypeScript MCP Server intended to be used with Claude Code

Verdict

BLOCK

Grade

Trust score

42 /100

Exposed tools

137 110r · 17w · 10d

Transport

stdio · streamable-http

License

—

Stars

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

[](https://mseep.ai/app/zachhandley-zmcptools)

[](https://opensource.org/licenses/MIT) [](https://www.typescriptlang.org/) [](https://nodejs.org/) [](https://modelcontextprotocol.io/)

🚀 TypeScript MCP Tools for Claude Code - Professional multi-agent orchestration platform with 61 enhanced tools, documentation intelligence, and advanced automation capabilities.

⚠️ Important Setup Note

Before spawning agents, run this command once to enable proper agent permissions:

claude --dangerously-skip-permissions

Agents run on daemon threads and need this permission to execute properly.

✨ Key Features

🎯 Multi-Agent Orchestration

Architect-Led Coordination: AI architect automatically spawns and coordinates specialized agent teams
Intelligent Dependencies: Agents work in proper order (Backend → Frontend → Testing → Documentation)
Real-Time Communication: Agents collaborate through dedicated chat rooms with message broadcasting
Foundation Session Caching: 85-90% cost reduction through automatic shared context management
Professional Task Management: Create, assign, track, and monitor complex development workflows

🎨 TypeScript-First Architecture

Type-Safe MCP Server: Built with Zod schemas and strict TypeScript for reliability
Modern CLI Interface: Commander.js-powered CLI with structured command hierarchy
Development Ready: One-command setup with hot-reload development via tsx
Binary Distribution: Global access via claude-mcp-tools and claude-mcp-server commands
Professional Build System: tsup-based compilatio

Read from source at commit 509dcbf1e3f1OBSERVED · 2026-10-08

Connect

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.

claude-code

claude mcp add zmcp-tools -- npx -y [email protected]

claude-desktop

{
  "mcpServers": {
    "zmcp-tools": {
      "command": "npx",
      "args": [
        "-y",
        "[email protected]"
      ]
    }
  }
}

Exposed tools (137)

110 read · 17 write · 10 destructive. Blast radius: 10 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.

Tool	Risk	Description
`Implementation`	read	Core feature implementation
`agent_types`	read	Preferred agent types for execution (comma-separated)
`analysis_type`	read	Type of analysis (technical, business, strategic, etc.)
`analyze_coordination_patterns`	read	Analyze coordination patterns and suggest improvements
`analyze_dom_structure`	read	AI-guided exploration and analysis of DOM structure using goal-oriented patterns. Analyzes stored DOM JSON to identify interactive elements, content areas, and navigation patterns.
`analyze_file_symbols`	read	Extract and analyze symbols (functions, classes, etc.) from code files
`analyze_project_structure`	read	Analyze project structure and generate a comprehensive overview
`analyze_screenshot`	read	AI-powered analysis of page screenshots with custom prompts. Can focus on specific regions and provide contextual insights.
`api-documenter`	read	Generate comprehensive API documentation with examples and schemas
`api_files`	read	API files or endpoints to document
`approach`	read	Preferred approach (investigation, implementation, testing)
`architect-orchestration`	read	Architect agent for complex multi-agent orchestration with sequential thinking
`architect-planning-template`	read	Template for architect agents to use sequential thinking for planning
`architecture`	read	Preferred architecture or patterns to use
`audience`	read	Target audience (developers, users, maintainers)
`backwards_compatible`	read	Require backwards compatibility (true/false)
`breakdown_depth`	read	Breakdown depth (shallow, medium, deep)
`broadcast_message_to_agents`	read	Broadcast a message to multiple agents with auto-resume functionality
`bug_description`	read	Description of the bug to fix
`cancel_scrape_job`	read	Cancel an active or pending scraping job
`cleanup_orphaned_projects`	read	Clean up orphaned or unused project directories
`cleanup_stale_agents`	read	Clean up stale agents with enhanced options and optional room cleanup
`cleanup_stale_analyses`	read	Clean up stale analysis files older than specified days
`cleanup_stale_rooms`	read	Clean up stale rooms based on activity and participant criteria
`close_browser_session`	read	[LEGACY] Close a browser session. Use manage_browser_sessions instead.
`close_room`	destructive	Close a communication room (soft delete - marks as closed but keeps data)
`code-review`	read	Comprehensive code review with security, performance, and quality analysis
`collection`	read	Vector collection to search (default: documentation)
`collection_name`	write	Vector collection name to create/manage
`complexity_level`	read	Complexity level (low, medium, high, very_high)
`constraints`	read	Constraints or requirements to maintain
`continue_agent_session`	read	Continue an agent session using stored conversation session ID with additional instructions
`create_browser_session`	write	Create a new browser session with intelligent auto-close and session management
`create_delayed_room`	write	Create a delayed room for coordination when agents realize they need it
`create_execution_plan`	write	Create comprehensive execution plan using sequential thinking before spawning agents
`create_knowledge_relationship`	write	Create a relationship between two entities in the knowledge graph
`create_task`	write	Create and assign task to agents with enhanced capabilities
`criteria`	read	Decision criteria or evaluation factors
`database_type`	read	Database type (sqlite, postgres, mysql)
`db-migration`	write	Plan and execute database schema migrations safely
`decision_context`	read	Context and background for the decision
`delete_all_website_pages`	destructive	Delete all pages for a website (useful for clean slate before re-scraping)
`delete_execution_plan`	destructive	Delete an execution plan by ID
`delete_pages_by_ids`	destructive	Delete specific pages by their IDs
`delete_pages_by_pattern`	destructive	Delete website pages matching URL patterns (useful for cleaning up version URLs, static assets)
`delete_room`	destructive	Permanently delete a communication room and all its messages
`depth`	read	Analysis depth (surface, detailed, comprehensive)
`documentation-writer`	write	Create comprehensive documentation for code, APIs, and features
`domain`	read	Problem domain (technical, business, architectural, etc.)
`easy_replace`	read	Fuzzy string replacement in files with smart matching
`embedding_provider`	read	Embedding provider (openai, default)
`execute_browser_script`	write	[LEGACY] Execute JavaScript in the browser context. Use interact_with_page instead.
`execute_with_plan`	write	Execute an objective using a pre-created execution plan with well-defined agent tasks
`feature-implementer`	read	Systematic feature implementation with planning and testing
`feature_description`	read	Detailed description of the feature to implement
`files_affected`	read	Comma-separated list of files that might be affected
`find_files`	read	Search for files by pattern with optional content matching
`find_related_entities`	read	Find related entities through relationship traversal
`force_unlock_job`	destructive	Force unlock a stuck scraping job - useful for debugging and recovery
`force_unlock_stuck_jobs`	destructive	Force unlock all stuck scraping jobs (jobs that haven\
`format`	read	Documentation format (markdown, API docs, inline comments)
`foundation_session`	read	Foundation session ID for cost reduction
`generate_project_summary`	read	Generate AI-optimized project overview and analysis
`get_cleanup_configuration`	read	Get current cleanup configuration and settings for agents and rooms
`get_execution_plan`	read	Retrieve a previously created execution plan
`get_page_screenshot`	read	Retrieve stored screenshot for a page. Returns file path or base64 encoded image data for AI visual analysis.
`get_project_overview`	read	Get comprehensive project overview from TreeSummary analysis
`get_scraping_status`	read	Get status of active and recent scraping jobs (worker runs automatically)
`include_examples`	read	Include request/response examples (true/false)
`integration_depth`	read	Integration depth (surface, detailed, comprehensive)
`interact_with_element`	read	[LEGACY] Interact with a page element. Use interact_with_page instead.
`interact_with_page`	read	Perform multiple interactions with a page: click, type, hover, select, screenshot, wait, scroll
`issue-fixer`	read	Single-agent focused problem solver for specific issues
`issue_description`	read	Description of the issue to resolve
`join_room`	read	Join communication room for coordination
`knowledge-graph-integration`	read	Integrate knowledge graph memory for better context and decision making
`knowledge_domains`	read	Specific knowledge domains to search (comma-separated)
`limit`	read	Maximum number of results (default: 10)
`list_agents`	read	Get list of active agents
`list_browser_sessions`	read	[LEGACY] List all browser sessions. Use manage_browser_sessions instead.
`list_documentation_sources`	read	List all configured documentation sources
`list_execution_plans`	read	List execution plans for discovery and monitoring
`list_files`	read	List files in a directory with smart ignore patterns
`list_room_messages`	read	List messages from a specific room with pagination
`list_rooms`	read	List communication rooms with filtering and pagination
`manage_browser_sessions`	read	Manage browser sessions: list, close, cleanup idle sessions, get status
`metrics`	read	Performance metrics to focus on (speed, memory, throughput)
`migration_type`	read	Type of migration (schema, data, index, cleanup)
`monitor_agents`	read	Monitor agents with real-time updates using EventBus system
`multi-agent-bug-fix`	read	Coordinate multiple agents to investigate and fix a bug with comprehensive testing
`navigate_and_scrape`	read	Navigate to a URL and optionally scrape content in one operation. Auto-creates session if needed.
`navigate_dom_path`	read	Navigate to specific elements in DOM JSON using dot notation paths (e.g.,
`navigate_to_url`	read	[LEGACY] Navigate to a URL in an existing browser session. Use navigate_and_scrape instead.
`objective`	read	Complex objective to decompose and plan
`options`	read	Available options or alternatives
`orchestrate_objective`	read	Spawn architect agent to coordinate multi-agent objective completion
`orchestrate_objective_structured`	write	Execute structured phased orchestration with intelligent model selection (Research → Plan → Execute → Monitor → Cleanup)
`performance-optimizer`	read	Analyze and optimize code performance with benchmarking
`planning_context`	read	Context requiring strategic planning
`priority`	read	Priority level (low, medium, high, critical)
`problem_description`	read	Complex problem to analyze and solve
`query`	read	Search query for semantic similarity matching
`remove_file_analysis`	destructive	Remove analysis data for a deleted file from the TreeSummary system
`report_progress`	read	Report progress updates for agent tasks and status changes
`repository_path`	read	Path to the repository
`requirements`	read	Specific requirements or acceptance criteria
`review_type`	read	Type of review (security, performance, quality, full)
`run_comprehensive_cleanup`	write	Run comprehensive cleanup for both agents and rooms with detailed reporting
`scrape_content`	read	[LEGACY] Scrape content from the current page. Use navigate_and_scrape instead.
`scrape_documentation`	read	Scrape documentation from a website using intelligent sub-agents. Jobs are queued and processed automatically by the background worker. Supports plain string selectors for content extraction.
`search_dom_elements`	read	Search for DOM elements by type, content, keywords, or attributes. Returns matching elements with their paths for further navigation.
`search_knowledge_graph`	read	Search the knowledge graph using semantic or basic search
`semantic-search`	read	Perform semantic search across documentation using vector embeddings
`send_message`	write	Send message to coordination room
`sequential-thinking-analysis`	read	Use sequential thinking for step-by-step analysis of complex topics
`sequential-thinking-architect`	read	Use sequential thinking for complex objective decomposition and planning
`sequential-thinking-decision`	read	Use sequential thinking for complex decision making processes
`sequential-thinking-problem-solver`	read	Use sequential thinking for complex problem analysis and solution development
`server_url`	read	ChromaDB server URL (default: localhost:8000)
`spawn_agent`	read	Spawn fully autonomous Claude agent with complete tool access
`stakeholders`	read	Key stakeholders and their requirements
`standards`	read	Coding standards or guidelines to follow
`store_knowledge_memory`	read	Store a knowledge graph memory with entity creation
`take_screenshot`	read	[LEGACY] Take a screenshot of the current page. Use interact_with_page instead.
`target`	read	What to document (API, feature, codebase, etc.)
`target_area`	read	Area to optimize (database, API, algorithms, etc.)
`target_files`	read	Files or patterns to review (e.g.,
`task-breakdown-specialist`	read	Specialized agent for hierarchical task breakdown and planning
`task_context`	read	Context of the task requiring knowledge integration
`terminate_agent`	destructive	Terminate one or more agents
`threshold`	read	Similarity threshold 0-1 (default: 0.7)
`topic`	read	Topic or subject to analyze
`update_execution_plan`	write	Update an execution plan\
`update_file_analysis`	write	Update or create analysis data for a specific file in the TreeSummary system
`update_project_metadata`	write	Update project metadata in the TreeSummary system
`vector-db-setup`	write	Set up and configure ChromaDB vector database for semantic search
`wait_for_messages`	read	Wait for messages in a room

Trust audit

BLOCKgrade F · trust 42/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.

Layer	What it checks	Result
L0	Provenance & inventory	PASS
L1	Static analysis of the code	FAIL
L2	Instruction surface (what it tells the agent)	FAIL
L3	Class-specific surface	WARN
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: declared (2 observation(s))
Network: declared (5 observation(s))
Shell: declared (8 observation(s))
Dependencies: not all pinned
Secrets in source: none-found

Findings (18)

HIGHCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

src/services/BrowserManager.ts:302

const dynamicFunction = new Function('document', 'window', `return (${selectorOrCode})`);

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table

HIGHCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

src/services/WebScrapingService.ts:995

const elements = await page.$$eval(selector, (elements) => {

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table

HIGHCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

src/tools/BrowserTools.ts:649

const func = new Function('...args', script);

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table