NavClaw

Smart driving — exhaustive route search, may outperform default navigation. 导航/自驾/极限避堵, dozens of routes. One-tap iOS/Android deep link. Supports 高德/Amap. 智能...

Security Scan

The skill's code, instructions, and required secrets line up with its stated purpose (Amap route planning and optional Mattermost delivery), but you should verify the upstream/source and be cautious about where you store API keys and bot tokens.

Purpose & Capability

NavClaw's described functionality (exhaustive routing using Amap, producing deep links, optional Mattermost posting) matches the code and runtime instructions. The Amap API key and optional Mattermost bot credentials are consistent with the stated features and are expected for the declared capabilities.

Instruction Scope

SKILL.md instructs the agent/operator to check memory for an Amap API key, prompt the user, and/or write the key (and optional Mattermost config) into config.py; it also instructs calling wrapper.py and possibly uploading logs or sending messages to Mattermost. These actions are in-scope for a navigation skill, but they involve writing/storing secrets and sending files/messages to an external chat server. The agent is also instructed to update OpenClaw long-term memory or MEMORY.md as an installation option, which grants persistent trigger capability if done.

Install Mechanism

There is no remote download/install step in the registry metadata; the package is instruction/code-bundle style with a small requirements.txt (requests). No third-party archive downloads or obscure URLs are used. Dependencies are minimal and proportional (only requests required).

Credentials

The registry lists no required env vars, and the skill expects users to populate API_KEY, MM_BASEURL, MM_BOT_TOKEN, and MM_CHANNEL_ID inside config.py. Those secrets are directly relevant: Amap key for routing queries and Mattermost token/URL/channel for message delivery. The number of credentials is appropriate for the feature set, but note they are stored in a config file (not environment variables).

Persistence & Privilege

The skill does not request 'always: true' and uses normal autonomous invocation. However, SKILL.md explicitly recommends adding triggers to OpenClaw long-term memory / MEMORY.md to make invocation convenient; doing so gives the skill persistent triggerability within your agent. That is expected for an agent skill but is a privileged persistence decision you should consciously allow or avoid.

Assessment

This package appears to do what it says: it queries Amap, generates route candidates, and can post results to Mattermost. Before installing: 1) Verify the source (the skill claims a GitHub repo but the registry metadata lists source/homepage as unknown — confirm the upstream repo and review it yourself). 2) Only provide an Amap API key and Mattermost bot token if you trust the code and host; prefer creating a bot account with minimal privileges and a dedicated Amap key. 3) Be aware the skill writes credentials into config.py (a plain file) and may add an entry to OpenClaw long-term memory; avoid storing secrets in long-term memory or broad-scope locations. 4) If you want minimal blast radius, run the tool locally (not letting the agent autonomously write config files), or inspect and run wrapper.py manually to control when messages/logs are uploaded. 5) If you need higher assurance, ask the publisher for an official homepage or signed release and/or perform a line-by-line audit of the bundled files (wrapper.py and navclaw.py) to confirm there are no unexpected network calls beyond Amap and configured Mattermost endpoints.

NavClaw 🦀 - 出行智能导航AI助手 / Smart Driving Navigator

概述 / Overview

智能避堵导航 — 极限搜索数十条路线，可能比官方导航更优。一键跳转手机导航APP（iOS/Android）。

Smart congestion-avoidance navigator — exhaustive search of dozens of routes, may outperform default navigation. One-tap deep link to mobile nav apps.

五阶段流水线规划（广撒网 → 精筛选 → 深加工 → 迭代优化 → 路线固化），导航 APP 通常只返回 2-3 条路线，NavClaw 短时间内探索数十种绕行组合。

5-phase pipeline (Wide Search → Fine Filter → Deep Processing → Iterative Optimization → Route Finalization). While navigation apps typically return 2-3 routes, NavClaw explores dozens of bypass combinations in seconds.

导航平台：目前支持高德，后续扩展更多平台。

Navigation platform: currently supports Amap (高德), more platforms coming soon.

---

⚠️ 前置条件 / Prerequisites：需要高德 Web 服务 API Key（免费申请）。获取优先级：

1. 从记忆中查找用户是否已提供过高德 API Key
2. 如果没有，询问用户是否有高德 API Key
3. 如果用户没有，引导获取：高德开放平台 → 控制台 → 创建应用 → 添加 Key（Web 服务）

拿到 Key 后填入 config.py 的 API_KEY 字段：

API_KEY = "你的高德API Key"

---

触发方式 / Trigger：

用户说"从 [起点] 到 [终点] 导航"、"导航从 [起点] 到 [终点]"或"navigate from [A] to [B]"即可自动执行。说"到家"时自动替换为 config.py 中的 DEFAULT_DEST。

Trigger when user says "navigate from [A] to [B]", "从 [起点] 到 [终点] 导航", or "drive to [destination]". Saying "到家" (go home) auto-substitutes DEFAULT_DEST from config.py.

---

工作流程 / Workflow：

调用 wrapper.py --origin "起点" --dest "终点"，经过五阶段规划（广撒网 → 精筛选 → 深加工 → 迭代优化 → 路线固化），生成大量路线方案（含绕行优化），自动发送 3 条消息到聊天平台 + 日志附件：

• 消息 1 — 完整对比表格
• 消息 2 — 快速导航链接
• 消息 3 — 最终推荐 + iOS/Android 一键深度链接

---

输出格式 / Output：

• Mattermost（原生支持，推荐优先使用）：

  需要先在 config.py 中配置 MM_BASEURL、MM_BOT_TOKEN、MM_CHANNEL_ID，然后直接运行 wrapper.py --origin "起点" --dest "终点"，自动发送 3 条消息 + 日志附件（优先采用Mattermost 消息和日志附件，如果不成功则用备份方案）。

Mattermost 发文件附件

OpenClaw Mattermost 插件不支持原生附件，用 curl 直调 API：

1. POST /api/v4/files 上传文件，拿 file_id
2. POST /api/v4/posts 发帖，带 file_ids 字段

• 其他聊天工具（备份方案）：

  运行 wrapper.py --origin "起点" --dest "终点" --no-send，结果输出到 stdout，OpenClaw 读取后转发给用户即可。

  OpenClaw 可读取 stdout 按 📨 消息 1/2/3 分段转发给用户。日志文件路径在末尾 📝 日志: log/navclaw/... 行中，不要发路径，要读取后发出来，如果不能发附件，给发原文内容。

  （一定要原样发给用户，各个消息，特别是链接要保留，不能舍弃）

强烈建议先用原生方法 / Native method recommended

---

安装配置 / Setup：

pip install requests → cp config_example.py config.py → 编辑填入高德 API Key、默认终点、Mattermost 配置（可选，包括MM_BASEURL，MM_BOT_TOKEN，MM_CHANNEL_ID，如果记忆或者配置没有，提示用户给出，如果用户没有就忽略。如果有，要写入config.py对应位置）。

---

文件位置 / Files：

• 调用入口：wrapper.py
• 核心引擎：navclaw.py
• 配置文件：config.py（需用户创建）
• 配置模板：config_example.py
• 日志目录：log/

---

聊天平台 / Chat Platforms：

目前内置支持 Mattermost（通过 wrapper.py），其他聊天工具 OpenClaw 帮我转发。

最简单的办法是直接聊天告诉 OpenClaw 运行并读取结果发送给你，支持任何聊天平台，稳定性和上下文长度取决于你的大模型 API。如果想节约 token、防止上下文截断、加快响应速度，可以自行扩展 wrapper.py 或让 OpenClaw AI 阅读现有 Mattermost 代码帮你适配新平台。

---

性能参考 / Performance：

• 短途无拥堵（迭代=0）：约 6 秒、15 次 API、10 条路线
• 长途有拥堵（迭代=1）：约 30 秒、150 次 API、40 条路线

首次使用建议 MAX_ITER = 0 验证配置正确，MAX_ITER = 1 深度优化可能找到比官方更快的路线。

---

依赖 / Dependencies：

• Python 3.8+
• requests（唯一第三方依赖）
• 高德 Web 服务 API Key

---

作者 / Author：

公益技能，免费开源。 / Community-driven, open-source, free for everyone.

• Email: [e-mail]
• 小红书 / Xiaohongshu: @深度连接
• GitHub: AI4MSE