MCPMaestro
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| HOST | No | Interface to bind. | 0.0.0.0 |
| PORT | No | HTTP port for the entire app (API + MCP + SPA). | 3000 |
| NODE_ENV | No | Log level toggle. | development |
| ADMIN_TOKEN | No | If set, all /api/* requests require Authorization: Bearer <token>. | |
| AGENT_SECRET | No | Key used to encrypt shared credentials (set a strong value in production). | change-me-in-production |
| MEMORY_MAX_MB | No | Default cap on the on-disk size of the knowledge-graph memory store (MB). A GUI/API override is persisted in the settings table and takes precedence. Writes refuse to exceed the budget. | 100 |
| DISABLED_TOOLS | No | Comma-separated global tool denylist enforced by centralized policy. | |
| ALLOWED_ORIGINS | No | Comma-separated origins allowed to call the API. Empty = allow all. | |
| JOB_CONCURRENCY | No | Number of background jobs that may run at once. | 2 |
| FS_ALLOWED_PATHS | No | Comma-separated authorized directories agents may read/write/search via the filesystem tools and run git in. Leave empty to disable those tools. | |
| DISCOVERY_ENABLED | No | false disables startup/scheduled discovery crawling. | true |
| DNS_ALLOWED_HOSTS | No | Comma-separated authorized names agents may query via DNS (wildcards like example.com.* supported). Leave empty to disable DNS tools. | |
| LLM_REGISTRY_URLS | No | Comma-separated registry URLs to seed LLM/AI model discovery. | |
| MCP_REGISTRY_URLS | No | Comma-separated registry URLs to seed MCP server discovery. | |
| SSH_ALLOWED_HOSTS | No | Comma-separated authorized hosts agents may SSH into. Leave empty to disable the SSH client. | |
| REQUEST_TIMEOUT_MS | No | Timeout for outbound HTTP fetches. | 10000 |
| TOOL_APPROVAL_TOKEN | No | Approval token required when tool approvals are enforced. | |
| RATE_LIMIT_WINDOW_MS | No | Window length for per-agent/per-tool/provider quotas. | 3600000 |
| DEFAULT_CACHE_TTL_SEC | No | Default TTL for cacheable tool responses. | 900 |
| DISCOVERY_INTERVAL_MS | No | How often the scheduled discovery runs. | 21600000 |
| WEBHOOK_SHARED_SECRET | No | Optional fallback secret for inbound /hooks/:webhookId calls. | |
| ENFORCE_TOOL_APPROVALS | No | If true, approval-marked tools require approvalToken. | false |
| RESEARCH_WORKER_ENABLED | No | Enable the autonomous research worker schedule and research.run tooling. | true |
| DEFAULT_CALLS_PER_WINDOW | No | Default per-agent quota window enforced by the centralized policy layer. | 120 |
| NETWORK_DISCOVER_TARGETS | No | Comma-separated authorized hosts/CIDRs to probe for device discovery, ping, traceroute, and port-scan. Leave empty to disable those tools. | |
| RESEARCH_WORKER_MAX_PAGES | No | Per-run cap on how many web-search result pages the worker expands. | 3 |
| RESEARCH_WORKER_MAX_TOKENS | No | Approximate per-run text budget before a non-forced worker run is skipped. | 12000 |
| RESEARCH_WORKER_INTERVAL_MS | No | Default cadence for the seeded idle-only research worker schedule. | 1800000 |
| RESEARCH_WORKER_MAX_TIME_MS | No | Per-request timeout budget for worker web extraction. | 20000 |
| SOCKET_ALLOWED_LISTEN_HOSTS | No | Comma-separated authorized bind hosts for socket.listen (for example 127.0.0.1,localhost). | |
| SOCKET_ALLOWED_LISTEN_PORTS | No | Comma-separated authorized listen ports or ranges for socket.listen (for example 46000-46010). | |
| RESEARCH_WORKER_ONLY_WHEN_IDLE | No | Require operator-idle status and no active MCP sessions before background research runs. | true |
| SOCKET_ALLOWED_CONNECT_TARGETS | No | Comma-separated authorized socket client hosts. Falls back to NETWORK_DISCOVER_TARGETS when unset. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/blindengineerMO/MCPMaestro'
If you have feedback or need assistance with the MCP directory API, please join our Discord server