隧道代理
tunelo
by claw-bot
Expose local services and files to the internet through a public HTTPS URL. Designed for AI agents — when you need to let a user preview files remotely, share a dev server, demo an app, or give temporary access to localhost. Use this whenever the user says "share this", "let me see it on my phone", "send me the link", or needs to access something running locally from another device/network.
安装
claude skill add --url github.com/openclaw/skills/tree/main/skills/claw-bot/tunelo文档
When to use tunelo
Use tunelo when the user needs to:
- Preview files remotely — "show me that PDF", "let me browse those files on my phone"
- Share a local dev server — "give me a link to your React app"
- Demo something to a colleague — "send me the URL so I can test it"
- Access localhost from another device — mobile testing, remote debugging
- Share a directory — project files, documents, media, datasets
Do NOT use tunelo when:
- The user only needs local access (use
python3 -m http.serverortunelo serve . --local) - The files are already hosted somewhere public
Install
curl -fsSL https://tunelo.net/install.sh | sh
If tunelo is not found after install, the binary is at /usr/local/bin/tunelo.
Commands
Expose a local HTTP service
tunelo http 3000 # Expose port 3000 → get public HTTPS URL
tunelo http 5173 # React/Vite dev server
tunelo http 8080 # Any local service
tunelo http 3000 --private # Require access code to visit
Serve files with web explorer
tunelo serve . # Current directory → public URL with file browser
tunelo serve ./dist # Specific directory
tunelo serve ~/Documents # Any path
tunelo serve . --local # Local-only preview (no tunnel, no public URL)
tunelo serve . -l -p 8000 # Local preview on port 8000
The file explorer runs in the browser — directory browsing, code syntax highlighting, markdown rendering, PDF viewer, image/video/audio playback, CSV/Excel tables. Everything is embedded in the binary, no dependencies.
Options
tunelo http <PORT> --relay my.server:4433 # Use a custom relay server
tunelo http <PORT> -H 192.168.1.100 # Forward to non-localhost
tunelo http <PORT> --private # Auto-generate access code
tunelo http <PORT> --code mysecret # Set specific access code
Default relay is tunelo.net:4433 (free public relay). Use --relay for self-hosted.
Typical agent workflows
User says "share these files with me"
tunelo serve /path/to/files
# Give the user the public URL from the output
User says "I want to see this on my phone"
# If there's a dev server running:
tunelo http 3000
# If it's just files:
tunelo serve .
User says "let my colleague test the API"
tunelo http 8080 --private
# Give them the Share URL (includes access code)
User says "preview this locally first"
tunelo serve ./dist --local
# Opens on http://localhost:3000, no public URL
How it works
Browser → HTTPS → Relay → QUIC tunnel → Client → localhost / file server
- Public HTTPS URL assigned automatically (random subdomain like
abc123.tunelo.net) - QUIC transport — encrypted, multiplexed, low latency
- Auto-reconnects if connection drops
- Session limit: tunnels expire after ~2 hours on the public relay
相关 Skills
Claude API
by anthropic
Build, debug, and optimize Claude API / Anthropic SDK apps. Apps built with this skill should include prompt caching. Also handles migrating existing Claude API code between Claude model versions (4.5 → 4.6, 4.6 → 4.7, retired-model replacements). TRIGGER when: code imports `anthropic`/`@anthropic-ai/sdk`; user asks for the Claude API, Anthropic SDK, or Managed Agents; user adds/modifies/tunes a Claude feature (caching, thinking, compaction, tool use, batch, files, citations, memory) or model (Opus/Sonnet/Haiku) in a file; questions about prompt caching / cache hit rate in an Anthropic SDK project. SKIP: file imports `openai`/other-provider SDK, filename like `*-openai.py`/`*-generic.py`, provider-neutral code, general programming/ML.
并行代理
by axelhu
Use when facing 2 or more independent tasks that can be worked on without shared state - dispatches parallel subagents using sessions_spawn for concurrent investigation and execution, adapted for OpenClaw
高光制作器
by bwbernardweston18
>