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

  • Claude Code CookbookPlugin
    • The `/pr-create`, `/pr-fix` and `/semantic-commit` commands call `git` and `gh`: they create commits and pull requests and edit descriptions and labels
    • The `/update-node-deps`, `/update-flutter-deps` and `/update-rust-deps` commands change a project's dependency files
    • `/screenshot` captures macOS screenshots and `/search-gemini` sends queries through the Gemini CLI
  • Claude Code CyberSecurity SkillsSkill
    • Includes offensive techniques: exploit development, payload generation and red team operations
    • Scripts run security tools and network commands on hosts
  • Claude Code GuideSkill
    • The set includes offensive security skills: recon, vulnerability testing and exploitation tooling
    • Some skills run scripts and external tools on the host
  • Claude Code My WorkflowWorkflow
    • Ships bypassPermissions mode by default with broad Bash(*), Edit(**) and Write(**) allows, prompting for almost nothing
    • Runs scripts, hooks and pre-commit gates that execute code and drive compilation and git
  • Claude Code RouterCLI
    • Stores model provider API keys
    • All agent traffic, including project code, goes through the gateway to the chosen providers
  • Claude Code SkillsPlugin
    • Skills deploy services and run commands in your environment
    • Some plugins reach external services and handle payments and keys
  • Claude Code Skills and AgentsPlugin
    • Skills run shell commands and external CLIs, access the network and work with GitHub
    • They need external API keys, and the manual install overwrites the ~/.claude directory
  • Claude Code VKontakte ChannelPlugin
    • Relays messenger input into an agent session that can run code on the machine
    • Requires running Claude Code with the unapproved channels flag
    • Stores a community token with messaging rights
  • Claude Code WeChat ChannelMCP server
    • Incoming WeChat messages become agent input and can trigger commands with its permissions
    • The WeChat sign-in token is stored as a plain file in the home directory
  • Claude Command SuitePlugin
    • Commands run git, builds, deployments and release rollbacks, and `remove-dead-code` deletes project files
    • Skills read Cloudflare and ElevenLabs keys from `.env`, work with Linear and send audio to the ElevenLabs cloud
    • The `/dev:watch` command and the `file-watcher` skill run `claude -p` on a file change event without your involvement
  • Claude Debugs For YouMCP server
    • Runs your project's program under the debugger and executes expressions in its context
    • Reads workspace files through `listFiles` and `getFileContent` and hands them to the agent
    • Listens on port 4711 with no access check and no explicit binding to a local address
  • Claude Dynamic Workflows on CodexSkill
    • Launches dozens of Codex agents that write to the project folder by default
    • Spends tokens on your Codex account, and you have to set the ceiling yourself
  • Claude ForgePlugin
    • The installer changes global Claude Code settings and hooks
    • Hooks run commands on every agent action
  • Claude Night MarketPlugin
    • The plugins read and edit repository files and run shell commands inside a Claude Code session
    • The hooks use gh credentials, post discussions and may open issues
    • Delegation through conjure sends file contents to external models
  • Claude SEOPlugin
    • The installer sets up a Python env and the Chromium browser
    • Stores Google API credentials in ~/.config/claude-seo
    • Makes network requests to audited sites
  • Claude SquadCLI
    • Agents can act without confirmation in --autoyes mode
    • Commits and pushes branches to the remote repository
  • Claude Trading SkillsSkill
    • Works with broker and market data APIs and keeps their keys in environment variables
    • Deals with real money and trading decisions, though it places no orders itself
  • Claude VideoSkill
    • The skill runs installers: Homebrew on macOS, uv, Python 3.12 and WhisperX packages, and a hook runs a settings check at every session start
    • In Gemini mode the whole clip, a local file included, is uploaded to Google, and cloud recognition sends audio to Groq or OpenAI
    • Gemini, Groq and OpenAI keys sit in ~/.config/watch/.env, and on explicit request the skill reads browser cookies or a cookie file
  • Claude Yandex SkillsSkill
    • The yandex-direct skill creates and changes real campaigns and bids, spending budget
    • The other skills read site and traffic data, but the tokens grant broad account access
  • Claude-Codex SkillsSkill
    • Delivery skills commit, push, publish releases and deploy
    • Implementation skills change project code and dependencies
  • Claude-MemPlugin
    • Hooks record agent actions and tool output
    • Runs a local worker with an HTTP API
    • Offers hosted memory storage by default
  • Claude-to-IMSkill
    • Runs a background daemon and stores bot tokens
    • Exposes agent control from an external chat, limited by allow lists
  • ClaudeDBGateSkill
    • Adds a 1C extension with an HTTP service that runs any 1C query language request
    • The skill calls the service through PowerShell and sees the base documents, catalogs and registers
    • A default token is hardcoded in the sources and must be changed
  • ClaudePrism: LaTeX editor with Claude CodeCLI
    • The agent inside the editor edits project files and runs commands, including Python in the project environment
    • The content of files Claude reads is sent to the Anthropic API
  • ClaudeRMCP server
    • The agent runs arbitrary R in your session, reads any file you can read and reaches the network
    • Third-party manuscripts, CSV files and HTML widgets may hide instructions aimed at the model, and the code runs at once
    • The command guard blocks only obvious calls and can be bypassed
  • Claw OrchestratorCLI
    • The runtime launches coding agents that edit files and run commands in the workspace
    • Contract checks run commands and HTTP requests on your behalf
    • Token spend goes against your own subscriptions and keys with several vendors, and the spend ceiling is set with the maxBudgetUsd parameter
  • ClawSecSkill
    • Changes persistent agent config and adds hooks
    • Runs scripts and can set up scheduled jobs
    • The self pen-testing package runs offensive checks
  • Clawd Cursor: desktop control through MCPMCP server
    • The agent gets control of a real desktop: it clicks, types and reads windows, including mail and messengers
    • Once granted, the consent and the Accessibility and Screen Recording permissions on macOS open access to every app on the machine
    • Launching with npx -y fetches the latest package version every time, so the author advises a global install for a tool like this
  • ClineCLI
    • The agent runs terminal commands and edits files
    • Messaging integrations use bot tokens
    • Code is sent to the chosen model provider
  • Clipcat SkillSkill
    • Posts videos to your TikTok account and can spend paid credits
    • Keeps the API key in a local config and sends uploaded photos and videos to a cloud service
    • Downloads TikTok and Douyin videos and installs the CLI with a curl | bash script
  • Cloud Run MCP serverMCP server
    • Deploys code to a Google Cloud production environment
    • Can create projects linked to a billing account
    • Makes services public by default
  • Cloud.ru Skills for AI AgentsSkill
    • Creates and removes real virtual machines and ML deployments billed by usage
    • Manages AI agents and MCP servers on the Evolution platform with your account's rights
  • CloudBase AI ToolkitPlugin
    • Manages the database, auth and functions in the vendor cloud
    • Deploys apps and backends with key-based access
  • CloudSword: cloud security testingMCP server
    • It needs access keys of a cloud account, and in MCP mode module parameters arrive as call arguments and can end up in the conversation with the model
    • Some modules change the account: they create RAM and CAM users, attach policies, issue access keys and deploy a cloud function
    • It is a dual-use tool: suitable for authorized assessments and equally for reconnaissance of accounts you do not own
  • Cloudflare MCP ServerMCP server
    • Manages live Cloudflare cloud infrastructure
    • Requires OAuth or an API token with resource permissions
    • Can change DNS, Workers, firewall and other products
  • Cloudflare MCP serversMCP server
    • Can modify Cloudflare account configuration
    • Has access to logs, DNS and security settings
  • CoCo Super IntelligenceSkill
    • The installer creates symlinks in `~/.claude`, `~/.cursor` and other editor directories and adds a rules block to `CLAUDE.md`
    • Hundreds of skills and commands can run commands and work on your project inside your session
    • The optional reverse-skill pack carries reverse-engineering and penetration-testing methods
  • Code ModeCLI
    • Runs code that calls external tools
    • Works with the network and OAuth tokens
  • CodeAFCLI
    • The agent edits repository files, runs tests and builds and merges checked work into your branch itself
    • It works with provider keys and by default sends numeric usage measurements to the model pool
    • The `do` mode and schedules run work without you and spend money on models
  • CodeAgentCLI
    • The agent itself runs shell commands and writes files, with approval needed only for dangerous commands and writes outside the allowlist
    • It talks to the model provider and to web pages, and the API key is stored in `~/.codeAgent/settings.json` in plain text
    • Hooks, cron jobs, background sub-agents and MCP connections run with the user's rights
  • CodeDBMCP server
    • The installer changes client configurations itself
    • Running a script from someone else's domain executes someone else's code
    • Reads and edits project files
  • CodePilot1C for 1C:EDTPlugin
    • The MCP Host can be configured to auto-allow all mutating tools with no confirmation
    • Access includes a runtime debugger and profiler that can affect configuration execution
  • CodeRunnerMCP server
    • Executes arbitrary agent code
    • By default sandbox code has network access
  • CodewhaleCLI
    • Edits project files and runs commands in the terminal
    • Installing by an internet script runs someone else's code
    • Requests go to the chosen model provider
  • Codex AutoresearchSkill
    • Autonomously changes code and creates and reverts git commits
    • Runs arbitrary metric commands and defaults to full access
  • Codex CLICLI
    • The agent runs commands and edits files
    • The installer downloads a binary
    • Code is sent to OpenAI models
  • Codex MCP Server: Codex inside Claude CodeMCP server
    • The codex tool runs the Codex CLI, and the fullAuto and workspace-write sandbox options let it edit project files
    • Code and questions go to the OpenAI API under your key, and the costs are billed to it
  • CodexProMCP server
    • Repository file editing goes through a public tunnel with a token in the address
    • The connection address must not be published
    • The server runs tools on your machine
  • Colab MCPMCP server
    • Runs code in your Colab session
    • The agent gets access to the session's files
  • Cold Outbound SkillsSkill
    • Cold mailings are regulated by the advertising law and 152-FZ
    • The skills spend money and limits through your API keys
    • Keys must not be published

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.