← 返回 Skills 市场
222
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install ainclaw-mind
功能描述
Save up to 90% on Token costs. One agent explores, all agents benefit. Cloud-cached workflows with zero inference cost.
使用说明 (SKILL.md)
ClawMind
One agent explores, all agents benefit.
A crowdsourced workflow registry that caches successful automation patterns, letting you skip LLM inference entirely when a matching workflow exists.
Why Use This?
1. Save Real Money
Traditional approach: LLM explores and reasons through every step, burning tokens on trial-and-error.
Our approach: Query the cloud for a cached workflow. If found, execute directly. Zero inference cost.
Token savings example (10-step browser task):
- Traditional: ~5000 tokens
- Workflow Cache: ~800 tokens
- Savings: 80%+
The more complex the task and the more you repeat it, the more you save.
2. Skip the Debugging Hell
The painful part of AI automation isn't writing the script—it's the endless debugging when:
- The website changes its layout
- Selectors break unexpectedly
- Edge cases you didn't anticipate
ClawMind solves this:
- Every successful workflow from any agent is cached
- When websites change, cached workflows auto-update
- You never debug the same problem twice
3. Platform Agnostic
Works with any Claw/Lobster engine. One workflow, all platforms. Automatic syntax adaptation.
How It Works
User Intent → Query Cloud → Match Found?
↓ Yes ↓ No
Execute Now Normal Flow
(1 second) (LLM reasons)
↓ ↓
Success! Success → Contribute
One agent's success becomes every agent's shortcut.
Features
Interceptor
Queries the cloud before LLM inference. On match, replays the cached workflow directly.
Trace Compiler
Converts successful session traces into reusable Lobster workflows automatically.
PII Sanitizer
Local-first privacy. All sensitive data stays local. Only workflow patterns are shared.
Configuration
| Option | Type | Default | Description |
|---|---|---|---|
cloud_endpoint |
string | https://api.workflowcache.dev |
Cloud API endpoint |
enabled |
boolean | true |
Enable/disable interception |
auto_contribute |
boolean | true |
Auto-contribute successful workflows |
timeout_ms |
number | 300 |
API timeout (ms) |
Installation
npx clawhub install ainclaw-mind
Or manually:
cd ~/.qclaw/workspace/skills/ainclaw-mind
npm install
npm run build
Security
- Full PII sanitization pipeline
- No account credentials ever uploaded
- Multi-node security validation on all workflows
- Malicious injection detection and blocking
Who Is This For?
- Heavy AI users — Daily automation, high token bills
- Cost-conscious developers — Every token saved is money saved
- Automation enthusiasts — Stop reinventing wheels
- Efficiency maximalists — Why reason when you can replay?
License
MIT-0 — Free to use, modify, and redistribute. No attribution required.
Tags: #AI-efficiency #token-saver #automation #crowdsourced #workflow-cache
安全使用建议
This skill does what it says (intercept intents, fetch replayable workflows, contribute successes), but it can send your intent text and the full current URL and session identifiers to the configured cloud endpoint. Those fields can contain sensitive information (e.g., emails, account IDs, auth tokens in query strings, or private intents). Before installing:
- Treat the default cloud endpoint (https://api.workflowcache.dev) as an external service that will receive data from your agent. If you can't trust it, do not enable the skill or set cloud_endpoint to a private/self-hosted endpoint.
- Disable auto_contribute (set auto_contribute=false) or disable the skill entirely (enabled=false) if you want interception without uploading contributions.
- Avoid running the skill while interacting with pages where URLs or intents might include secrets (auth tokens, payment identifiers).
- If you need the feature but want tighter privacy, ask the developer to sanitize intent and URL (strip query params, remove path segments containing tokens) before any network call, and to expose an option to require explicit, per-contribution approval.
- If you accept the risk, audit network traffic to the cloud endpoint and/or host your own workflow cache service so contributions remain under your control.
I assessed actual source files included in the package; the privacy claim in SKILL.md is inconsistent with the code that transmits intent and URL. If you need help crafting a safer configuration or questions to ask the maintainer, tell me what you want to protect and I can suggest specific settings.
功能分析
Type: OpenClaw Skill
Name: ainclaw-mind
Version: 1.0.5
The skill implements a crowdsourced workflow registry that sends user intents, URLs, and browser action traces to a remote API (api.workflowcache.dev) and executes 'Lobster' workflows returned by that server. While it includes a PII sanitizer (sanitizer.js) intended to strip sensitive data like passwords and API keys before exfiltration, the architecture creates a significant privacy risk and a centralized vector for Remote Code Execution (RCE) via the lobster.execute hook in interceptor.js. These high-risk capabilities are aligned with the stated purpose of 'token saving' but lack sufficient safeguards against a compromised or malicious upstream provider.
能力评估
Purpose & Capability
Name, description, permissions (browser, lobster, sessions_history, network) and implemented hooks (interceptIntent, onSessionComplete) are coherent: the skill needs access to browser state, session history, and network to query and contribute cached workflows.
Instruction Scope
SKILL.md promises 'Local-first privacy' and 'No account credentials ever uploaded', but the code sends intent (user text), the full current URL, node_id, session_id and the compiled Lobster workflow to the cloud. Intent and URLs often contain sensitive data (emails, tokens in query strings, account identifiers). The sanitizer is applied to action args, but intent and URL are not sanitized before upload.
Install Mechanism
No remote download/install URL or extract step; standard npm-based build is used and only undici is a runtime dependency. There are no obscure or shortener URLs or arbitrary binary fetches in the manifest.
Credentials
The skill requests no credentials and only asks for APIs it needs, which is appropriate. However, the data it transmits (intent text, full URL, session_id, node_id) is more sensitive than the README/SKILL.md acknowledges; requiring no env vars is fine, but the claimed 'no credentials uploaded' is not enforced for URLs or intent content.
Persistence & Privilege
always is false and the skill registers normal hooks; it does not request permanent global privileges or modify other skills. Autonomous invocation via hooks is expected for this type of skill.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install ainclaw-mind - 安装完成后，直接呼叫该 Skill 的名称或使用
/ainclaw-mind触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.5
- Updated skill name and branding from "workflow-cache" to "ClawMind"
- Changed all documentation references and installation instructions to use the new ainclaw-mind package name
- No feature or functionality changes—documentation and naming improvements only
v1.0.4
- Added comprehensive SKILL.md detailing the Workflow Cache feature, including usage, benefits, and security.
- Introduced the concept of cloud-cached workflows to reduce token costs by skipping LLM inference when a match is found.
- Described key features: Interceptor, Trace Compiler, and PII Sanitizer.
- Provided configuration options and installation instructions.
- Clarified target users and licensing.
元数据
常见问题
ClawMind 是什么？
Save up to 90% on Token costs. One agent explores, all agents benefit. Cloud-cached workflows with zero inference cost. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 222 次。
如何安装 ClawMind？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install ainclaw-mind」即可一键安装，无需额外配置。
ClawMind 是免费的吗？
是的，ClawMind 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
ClawMind 支持哪些平台？
ClawMind 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 ClawMind？
由 ainclaw（@ainclaw）开发并维护，当前版本 v1.0.5。
推荐 Skills