技能详情（站内镜像，无评论）
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 0 · 146 · 0 current installs · 0 all-time installs
⭐ 0
安装量（当前） 0
🛡 VirusTotal ：良性 · OpenClaw ：良性
Package：amandaclarke61/toybridge
安全扫描（ClawHub）
- VirusTotal ：良性
- OpenClaw ：良性
OpenClaw 评估
The skill's instructions, requirements, and lack of installs/credentials are internally consistent with its stated purpose of controlling locally running ToyBridge-connected BLE toys.
目的
The name/description claim to control BLE toys via a locally running ToyBridge server and the SKILL.md only requires that server; no unrelated credentials, binaries, or installs are requested. Minor oddities: the registry metadata restricts the skill to macOS (darwin) even though the ToyBridge server appears OS-agnostic, and the README refers to specific repo paths (4-bridge/ble_worker.py) which are part of the ToyBridge project rather than th…
说明范围
Runtime instructions are explicit: use bash/curl to POST to local endpoints (host.docker.internal:8888 or localhost). The skill does not instruct reading arbitrary system files or sending data to external endpoints. It references local ToyBridge files and a server start command only as prerequisites. One small vagueness: the command 'uv run 4-bridge/server.py' is nonstandard/unexplained (likely a project-specific runner) but not a security red…
安装机制
No install spec or code files are included; this is instruction-only, so nothing will be downloaded or written to disk by the skill itself.
证书
The skill requests no environment variables, no credentials, and no config paths. The local-only HTTP API calls are proportional to the described device-control purpose.
持久
always is false (normal). The skill allows autonomous invocation (disable-model-invocation is false) which is the platform default. Because the skill can issue commands that operate physical devices, you should consider whether you want an agent to be able to call it autonomously — this is a policy/consent question rather than an incoherence in the skill.
综合结论
This skill is internally coherent: it simply issues curl requests to a local ToyBridge server and requires no installs or secrets. Before installing, make sure you actually run the ToyBridge server locally and review that upstream repo (https://github.com/AmandaClarke61/toybridge) to ensure you trust the code that speaks to your device. Be aware that the agent will send HTTP requests to localhost to control physical hardware — if you do not wa…
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「ToyBridge」。简介：Control any BLE toy that has been reverse-engineered and connected via the ToyB…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/amandaclarke61/toybridge/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
---
name: toybridge
description: Control any BLE toy that has been reverse-engineered and connected via the ToyBridge server. Calls a local HTTP API to send vibrate/stop commands. Requires the ToyBridge server running on the same machine.
metadata: {"openclaw": {"os": ["darwin"]}}
---
# ToyBridge — Universal BLE Toy Control
Control **any BLE toy** through OpenClaw, as long as you have the [ToyBridge](https://github.com/AmandaClarke61/toybridge) server running.
This skill is for devices that are **not supported by Buttplug.io/Intiface** — devices with proprietary or unknown protocols that you've reverse-engineered yourself using the ToyBridge toolkit.
> If your device IS supported by Buttplug.io, use the `intiface-control` skill instead — it's easier.
---
## Prerequisites
1. You've reverse-engineered your device's BLE protocol using [ToyBridge](https://github.com/AmandaClarke61/toybridge)
2. You've configured `4-bridge/ble_worker.py` for your device
3. The ToyBridge server is running: `uv run 4-bridge/server.py`
See the [full setup guide](https://github.com/AmandaClarke61/toybridge) for step-by-step instructions.
---
## Commands the agent will use
### Vibrate at intensity
```bash
curl -s -X POST http://host.docker.internal:8888/vibrate
-H "Content-Type: application/json"
-d '{"intensity": 60}'
```
`intensity`: 0–100 (0 = stop)
### Stop immediately
```bash
curl -s -X POST http://host.docker.internal:8888/stop
```
### Check status
```bash
curl -s http://host.docker.internal:8888/status
```
> If OpenClaw runs natively (not in Docker), replace `host.docker.internal` with `localhost`.
---
## Intensity guide
| Range | Feel |
|--------|---------|
| 1–20 | Gentle |
| 30–50 | Medium |
| 60–80 | Strong |
| 90–100 | Maximum |
---
## Preset patterns
| Pattern | What it does |
|---------|-------------|
| `pulse` | Bursts of 80%, 5 times |
| `wave` | Ramp up 20→100%, then back down, x2 |
| `tease` | 30% → 70% → 100%, escalating, then stop |
To run a pattern:
```bash
curl -s -X POST http://host.docker.internal:8888/vibrate
-H "Content-Type: application/json"
-d '{"pattern": "wave"}'
```
---
## Agent rules
- Always stop (intensity 0) after a timed session unless user says to keep going
- Do **not** use the `notify` tool — use `bash` with `curl`
- Replace `host.docker.internal` with `localhost` if OpenClaw is not in Docker
---
## Troubleshooting
| Problem | Fix |
|---------|-----|
| `connection refused` | Make sure `uv run 4-bridge/server.py` is running |
| Device doesn't respond | Check your device config in `ble_worker.py` |
| Wrong intensity | Values are clamped to 0–100 |