SkillFoxx

Type a name, a task or a service

Match with Foxx AI

Access levels of skills and MCP servers

The access level of every catalog entry and the reasons behind it. The level shows what a tool adds to what the agent can already do. It does not rate code quality.

Found: 1417

  • Kommo MCP server with token auto-refreshMCP server
    • Tools create and change deals, contacts, companies and tasks in a real account
    • Access and refresh tokens are stored in the MCP client configuration
  • Kommo MCP: 152 tools for Kommo CRMMCP server
    • 152 tools cover writing to almost all Kommo entities, including webhooks and salesbots
    • OAuth tokens with auto-refresh are saved in the .kommo_tokens.json file
  • Kontur Elba MCP serverMCP server
    • Creates bills, acts and delivery notes that go out to counterparties
    • Changes product and counterparty records in real accounting data
  • Kontur.Diadoc MCPMCP server
    • Sends and signs legally binding documents
    • Stores the Diadoc account login and password
  • Kontur.Elba MCP serverMCP server
    • The elba_request tool can create and change invoices, waybills, UPD and bank accounts
    • Any POST, PUT or DELETE in the Elba schema changes real accounting data
  • KubbPlugin
    • The commands and MCP server run `kubb.config.ts` as code, and `kubb init` installs `@kubb/plugin-*` packages from npm
    • Generation writes files into the project and, with `output.clean: true`, wipes the output folder before every run
    • A spec can be pulled from a URL, and the skill advises validating untrusted ones with `kubb validate`
  • Kubernetes MCPMCP server
    • Acts with the permissions of the current kubectl context
    • Can delete resources, drain nodes and run arbitrary kubectl commands
    • Logs may contain sensitive data
  • Kubernetes MCP ServerMCP server
    • Can modify and delete cluster resources and exec into pods
    • Uses kubeconfig credentials
  • L HarnessMCP server
    • The agent sends messages and broadcasts to real customers
    • It sees customers' profiles, chats and form answers
    • The API key gives access to staff and ad platform management in the account
  • LINE Bot MCP ServerMCP server
    • Sends messages to real users from your account
    • Broadcasts to all followers of an official account
    • Handles a channel access token and manages rich menus
  • Lamoda seller cabinet MCP serverMCP server
    • Write methods in the catalog change prices, stock and orders in the live seller account
    • Response and parameter shapes are unverified against the live API, raising the risk of mistakes on first use
    • Irreversible operations require two confirmations, but the user is still responsible for checking them
  • Lanshu AI Presenter VideoSkill
    • An authorized portrait is needed: someone else's likeness cannot be used without consent
    • The finished video must be marked as synthetic
    • The skill calls paid services with your keys
  • Lark CLICLI
    • The agent acts in your name: it sends messages and mail, edits docs and tables, approves and rejects approvals
    • After login the CLI keeps authorization data in the system keychain and reads chats, mail and docs within the granted scope
    • Installation runs a `postinstall` script, and service signals about the system and device go to Feishu and Lark
  • LateCLI
    • Subagents run shell commands and edit project files
    • The i-promise-i-have-backups-and-will-not-file-issues mode runs every tool without confirmation
    • Cloud model keys sit in environment variables and config.json
  • LaunchDarkly MCP ServerMCP server
    • Changes and deletes flags that affect production immediately
    • The API key is stored in the client config
  • LayrrCLI
    • Hands the agent an edit task, and the agent changes project files
    • Creates a git repository if there is none and commits edits into your history
    • Proxies your dev server traffic through local port 4567
  • Lemma skillsSkill
    • The skills drive a working Lemma account through the command line: they create and change tables, workflows and apps and import them into the space
    • The agents you build work with external services through connectors, for example Gmail and Slack
    • The installer overwrites existing `lemma-*` skills in your agent's skills directory
  • LetsFGMCP server
    • Books real flights and hotels and holds payment on the connected card
    • Requires real passenger passport and contact details
    • Talks to the external commercial letsfg.co service and its sellers
  • LifeOSSkill
    • Installation writes hooks into the agent settings.json and places files under ~/.claude, and those hooks later run during your sessions
    • The quick command pipes a script from the network into bash
    • The system accumulates personal context: goals, contacts, notes and memory of past sessions
  • Linear CLICLI
    • A Linear API key gives read and write access to the workspace: issues, comments, projects, documents
    • The CLI can delete and archive issues, milestones and documents
    • The --public flag publishes an image at a link without authentication, and the skill lets the agent call curl
  • LinghunCLI
    • Runs shell commands, edits project files and creates git checkpoints on the user's machine
    • Stores the model API key locally in a user-level provider.env file
  • LinkedIn Marketing SkillsSkill
    • Publishes posts and comments to LinkedIn through Publora after approval
    • Reads LinkedIn data via third-party Apify actors and stores API keys for paid services
  • Local REST API for ObsidianMCP server
    • The API key allows writing and deleting any vault file and running any Obsidian command
    • Notes and attachments the agent reads through MCP enter the model context and go to its provider
    • The server listens on 127.0.0.1 only, but any local program that knows the key can call it
  • Loki ModeCLI
    • The agent runs commands, writes code and commits autonomously
    • Skill mode requires skipping permission prompts
    • Consumes paid model tokens
  • LongHorizon-HarnessCLI
    • It drives an agent that runs commands and edits files for hours on end
    • It controls the desktop and terminal with real actions
    • It acts in the launch directory, that is your real project
  • LoonflowMCP server
    • The agent can act on tickets: approve, reject and change process data
    • A personal access token gives the agent the same ticket and user-list rights as your account
    • The tenant's user list contains employee names, emails and departments
  • Loop EngineeringPlugin
    • At higher autonomy levels it makes edits and opens PRs without line-by-line human review
    • Regular unattended runs can noticeably increase token spend
  • MATLAB MCP ServerMCP server
    • Runs arbitrary MATLAB code on the user's machine
    • Starts and quits MATLAB processes and works with project files
  • MAX AI secretary guideWorkflow
    • Uses MAX's unofficial User API via a secondary SIM, not the official Bot API
    • Misuse (mass mailing, automating on behalf of other people) falls under 152-FZ, 38-FZ and Criminal Code art. 159, which the README warns about explicitly
  • MAX MCPMCP server
    • Sends messages and files as your account
    • Uses an unofficial client, so the account may be restricted
    • A session with access to all chats is stored locally
  • MAX channel for Claude CodePlugin
    • Gives direct access to a live Claude Code session through MAX messages
    • The /restart command can restart the session, and voice transcription is sent to the third-party Groq service
  • MAX channel for Claude CodePlugin
    • Messenger messages drive an agent session with access to your machine
    • Launch requires a flag for unverified development channels
    • The bot token is stored in ~/.claude/channels/max/.env
  • MAX messenger MCP serverMCP server
    • Sends, edits and deletes messages on behalf of the bot
    • Reads chat history and member lists
  • MCD Saver StrategistSkill
    • The McDonald's token opens your account: the agent claims every available coupon with `auto-bind-coupons`
    • The order is created from your account and leads to payment through a link
  • MCP Adapter for WordPressMCP server
    • Performs site actions with the chosen user's rights, up to administrator
    • Exposes an HTTP endpoint on the site
  • MCP AtlassianMCP server
    • Creates and edits issues and pages in an external system
    • Holds API tokens with user-level rights
  • MCP Claude CodeMCP server
    • The `write`, `edit` and `run_command` tools change files and execute commands in the accessible directories
    • The server reads every file in the allowed directories, including files with secrets
    • With sub-agents enabled, task text goes to the model provider you chose
  • MCP Code Execution ServerMCP server
    • The agent executes arbitrary Python, even if inside a container with no network and no root
    • Through the code proxies every MCP server from your configurations is reachable, including those that write data and use keys
  • MCP Dev LATAMMCP server
    • Payment and banking API keys give the agent access to accounts, charges and statements
    • The servers create payments, issue tax documents and send customers messages
    • Packages for many third-party providers install from npm and run through npx
  • MCP DigitalOcean IntegrationMCP server
    • Creates and deletes cloud resources, which affects billing
    • Works with production infrastructure and databases
  • MCP GatewayMCP server
    • Deploys infrastructure and manages access
    • Protocol version 2026-07-28 is needed
    • For platform teams
  • MCP Gateway RegistryMCP server
    • Manages access to a company's servers and agents
    • A community project and not a vendor's
  • MCP HubMCP server
    • Runs any command from the config, including through the `${cmd: ...}` substitution
    • Holds the tokens and keys of every server and passes them a shared environment through `MCP_HUB_ENV`
    • Calls the tools of any connected server through the REST API, and no access check for the hub itself is described
  • MCP Kali ServerMCP server
    • Runs arbitrary commands in a Kali terminal
    • Built for offensive security; use only on your own systems and with authorization
    • A server exposed on an external interface grants remote shell access
  • MCP LinearMCP server
    • Creates and changes issues, projects and webhooks in an external service
    • Can issue tokens and rotate secrets
  • MCP MongoDB Server (kiliczsh)MCP server
    • Without --read-only, by default it gives an agent insert, update and index-creation access to a real MongoDB database
    • The connection string carries database credentials and is passed as a command argument or environment variable
    • The --allow-cross-db and --allow-server-js flags remove the guards against reaching outside the database and running arbitrary code on the MongoDB server
  • MCP SSH ManagerMCP server
    • A remote shell on your servers
    • The default mode has no limits
    • The GitHub version is a beta
  • MCP Server Playwright by Automata LabsMCP server
    • Runs arbitrary JavaScript on opened pages
    • Can fill forms and click buttons on any site
    • Downloads and runs Playwright browsers
  • MCP Server TauriMCP server
    • The agent runs JavaScript in your app's webview, types text and calls IPC commands
    • Screenshots, DOM and logs of the app land in the model context
  • MCP Server TrelloMCP server
    • A Trello token gives the agent write access: it creates, modifies and archives cards and attaches files
    • The key and token sit in environment variables and the client config
    • The chosen board and workspace persist in ~/.trello-mcp/config.json

How we set the access level

Methodology

The editors set the level and the reasons from the README, the way the tool runs and the permissions it needs. Reading project files the agent can already see does not raise the level. There is no automatic code audit yet.