- Conduid
- Marketplace
- #permissions
MCP servers tagged permissions
31 live MCP servers tagged "permissions", ranked by trust score. Tags come from package metadata and repository topics, so this list covers servers across every category that work with permissions.
Wassette
Wassette: A security-oriented runtime that runs WebAssembly Components via MCP
Aci
ACI.dev is the open source tool-calling platform that hooks up 600+ tools into any agentic IDE or custom AI agent through direct function calling or a unified MCP server. The birt…
access
Access, a centralized portal for employees to transparently discover, request, and manage their access for all internal systems needed to do their jobs
Theauth
Auth for AI agents and humans. First-class agent identity, MCP OAuth 2.1, delegation, audit. TypeScript, edge-native, MIT.
Toolpermit
A local-first permission firewall and approval layer for AI agent tool calls.
Permit MCP Server
MCP Server for Permit.io read-only operations
Vaultagent MCP Proxy
MCP proxy server with VaultAgent permission controls — adds policy enforcement and audit logging to any MCP server
Devutils
MCP server with 45 developer utilities (16 free + 29 pro). Free: UUID, hash, HMAC, base64, timestamps, JWT decode, random strings, URL encode/decode, JSON format, YAML/JSON conver…
Browser Link
MCP server that bridges Claude Code, OpenCode, GitHub Copilot CLI and other MCP clients to a Chrome tab. Per-tool permissions, multi-agent mode (multiple MCP clients sharing one b…
Agentsid
Identity, permissions, and audit for AI agents. The Auth0 for the agent economy.
Security With Permissions And Elicitation
Secure MCP (Model Context Protocol) server with role-based permission management, elicitation-based user approval workflows, and full audit logging — built with FastMCP, Python &…
Unify
Unify MCP servers behind one endpoint with lazy loading, Python plugins, and role-based filtering for lower overhead and simpler config
Toolkit Examples
🔧 Build safe and reliable tools for Model Context Protocol (MCP) servers to enhance AI agent interactions with external systems.
Monoid Agent Kernel
Embeddable agent kernel for durable, provider-neutral LLM agents with permissions, pluggable tools, observability, and subagents.
Authz Spicedb Demo
AI agent tool-call authorization using SpiceDB (Google Zanzibar) via Model Context Protocol (MCP)
Claude Code Deepseek 3p Parity Skill
修复 Claude Code Desktop 接入 DeepSeek/第三方模型后能力缺失、联网失败、频繁权限弹窗的 Claude Skill
Capmark
Capability manifests for AI agent plugins. Declare what a plugin may do, in Markdown, and check it.
KavachOS
Auth for AI agents and humans. First-class agent identity, MCP OAuth 2.1, delegation, audit. TypeScript, edge-native, MIT.
Agentperms
Least-privilege permissions for AI agents and MCP tools. Record what your agent does, auto-generate a safe policy, and block dangerous tool calls before they happen.
YakAgent
Lightweight Python agent framework — tool-calling loop, permissions, sub-agents, streaming. 17 LLM providers, 15 built-in tools.
Agent Permission Graph
Map, score, and explain what local AI agents and MCP servers are allowed to touch.
Query
Secure local MCP server for Claude Code — query databases with Apple Keychain credentials, granular SQL permissions, and full audit logging.
Permission Linter
CLI and GitHub Action-ready linter for MCP permissions, secrets, destructive tools, and least-privilege policy.
Permission Policy Tester
Scan MCP tool descriptions and schemas for risky permission policy signals.
Agent Capability Middleware
Open SDK and reference profiles for user-controlled AI agent permissions, private context, and x402 policy flows
Coffee Desk MCP
Demo mcp server for coffeedesk.app (mcp.coffeedesk.app)
Toolhutch
Local-first permission auditor for agent tool manifests.
Tool Safety Lab
Public-safe MCP-style tool safety lab with permission gates, approval boundaries, schemas, audit logs, and CI tests.
Shell Guard
MCP server for safe shell access. Every command validated against per-agent permission rules before execution. Powered by AgentsID.
Claude MCP Allow
Generate least-privilege Claude Code permission rules for MCP tools from their live annotations: readOnlyHint tools go to allow, everything else to ask, one mcp__server__tool rule…
Incident Agent Claude Code
No description published.