← 返回 Skills 市场

Aiprox Workflows

Name: Aiprox Workflows
Author: unixlamadev-spec

作者 unixlamadev-spec · GitHub ↗ · v1.0.1 · MIT-0

cross-platform ✓ 安全检测通过

248

总下载

当前安装

版本数

在 OpenClaw 中安装

/install aiprox-workflows

功能描述

Create and run multi-agent AI workflows on AIProx. Chain agents into scheduled pipelines. Pay per execution in sats.

使用说明 (SKILL.md)

\r \r

AIProx Workflows\r

\r AIProx Workflows is a multi-agent pipeline engine. Chain AI agents into named workflows, schedule them to run on a cron, and get results delivered by email or webhook. Pay per execution in sats — no subscriptions, no monthly fees.\r \r

When to Use\r

Building automated pipelines that chain multiple AI agents\r
Scheduling recurring tasks (daily news digest, competitive intel, market signals)\r
Running one-off multi-step pipelines on demand\r
Any workflow where one agent's output feeds into the next\r \r

MCP Tools\r

`create_workflow`\r

\r Create a named pipeline with ordered agent steps.\r \r

Create a workflow "daily-crypto-brief" that searches Bitcoin news, analyzes sentiment, and emails a summary @daily\r
```\r
\r
Step chaining: use `$step1.result`, `$step2.result`, etc. to pass outputs forward.\r
\r
**Available capabilities:**\r
\r
| Capability | What it does |\r
|---|---|\r
| `web-search` | Real-time web search via search-bot |\r
| `sentiment-analysis` | Sentiment and tone analysis via sentiment-bot |\r
| `scraping` | Web scraping and article extraction via data-spider |\r
| `data-analysis` | Data processing and text analysis via doc-miner |\r
| `translation` | Multilingual translation via polyglot |\r
| `vision` | Image/screenshot analysis via vision-bot |\r
| `code-execution` | Code audit and security review via code-auditor |\r
| `email` | Send email notifications via email-bot |\r
| `market-data` | Prediction market signals via market-oracle |\r
| `token-analysis` | Solana token safety and rug detection via isitarug |\r
\r
### `run_workflow`\r
\r
Trigger any workflow by its ID.\r
\r
```\r
Run workflow wf_abc123\r
```\r
\r
### `list_workflows`\r
\r
Show all workflows for the current spend token.\r
\r
```\r
List my workflows\r
```\r
\r
### `get_run_history`\r
\r
Show past execution results, sats spent, and step-by-step output.\r
\r
```\r
Show run history for wf_abc123\r
```\r
\r
### `delete_workflow`\r
\r
Delete a workflow (cancels scheduled runs).\r
\r
### `run_template`\r
\r
Run a pre-built template by name in one shot.\r
\r
```\r
Run the polymarket-signals template and email me at [email protected]\r
```\r
\r
## Pre-Built Templates\r
\r
| Template name | Pipeline | Cost |\r
|---|---|---|\r
| `news-digest` | search-bot → sentiment-bot → email-bot | ~150 sats/run |\r
| `token-scanner` | data-spider → isitarug → email-bot | ~120 sats/run |\r
| `competitive-intel` | search-bot → doc-miner → sentiment-bot → email-bot | ~200 sats/run |\r
| `multilingual-content` | data-spider → doc-miner → polyglot | ~180 sats/run |\r
| `site-audit` | vision-bot → code-auditor → doc-miner | ~220 sats/run |\r
| `polymarket-signals` | market-oracle → sentiment-bot → email-bot | ~160 sats/run |\r
\r
## Authentication\r
\r
Set `AIPROX_SPEND_TOKEN` in your MCP server config. Get a spend token at [lightningprox.com](https://lightningprox.com).\r
\r
```json\r
{\r
  "mcpServers": {\r
    "aiprox-workflows": {\r
      "command": "npx",\r
      "args": ["aiprox-workflows-mcp"],\r
      "env": {\r
        "AIPROX_SPEND_TOKEN": "lnpx_your_token_here"\r
      }\r
    }\r
  }\r
}\r
```\r
\r
## Scheduling\r
\r
| Shorthand | When |\r
|---|---|\r
| `@hourly` | Every hour |\r
| `@daily` | Every day at midnight |\r
| `@weekly` | Every Sunday |\r
| `0 9 * * 1-5` | 9am Monday–Friday |\r
\r
## Pricing\r
\r
50–220 sats per workflow execution depending on agents used. No monthly plan. Deducted from your Lightning spend token balance.\r
\r
## Behavior Guidelines\r
\r
**When a user asks to automate something:**\r
1. Identify which capabilities are needed (search → analyze → notify is the common pattern)\r
2. Suggest a template if one matches, otherwise offer to create a custom workflow\r
3. Confirm the schedule and email/webhook if they want recurring runs\r
4. After creating, offer to run it immediately\r
\r
**Cost transparency:**\r
- Mention estimated sats cost before creating/running\r
- Report `sats_spent` after each run\r
\r
**Check templates first:**\r
If the task matches a template, use `run_template` — it's faster and pre-configured.\r
\r
## Links\r
\r
- Dashboard: [aiprox.dev/workflows](https://aiprox.dev/workflows)\r
- Templates: [aiprox.dev/templates](https://aiprox.dev/templates)\r
- Registry: [aiprox.dev/registry.html](https://aiprox.dev/registry.html)\r
- npm SDK: `npm install aiprox-workflows`\r
- MCP server: `npx aiprox-workflows-mcp`\r

安全使用建议

This skill is coherent with its stated purpose, but before installing consider: (1) Treat AIPROX_SPEND_TOKEN as a sensitive spend-only credential; ask whether the token can be scoped/limited and whether revocation is easy. (2) Confirm whether the agent will run workflows autonomously by default and what user consent/confirmation is required before each paid run — if you want to avoid unexpected charges, disable autonomous invocation or require explicit approval. (3) Verify the external domains mentioned (aiprox.dev, lightningprox.com) and the npm package (aiprox-workflows) before installing any adapter. (4) Start by creating/running a single test workflow and monitor run_history and sats_spent; set strict budget caps if the service supports them. (5) If you need stronger guarantees, ask the provider how emails/webhooks are sent/stored and whether sensitive data processed in workflows is retained or accessible to third parties.

功能分析

Type: OpenClaw Skill Name: aiprox-workflows Version: 1.0.1 The aiprox-workflows skill bundle provides a legitimate interface for managing multi-agent AI pipelines via the AIProx service. It includes tools for creating, running, and monitoring workflows that utilize capabilities like web searching, scraping, and sentiment analysis, with payments handled via a Lightning Network spend token (AIPROX_SPEND_TOKEN). The SKILL.md instructions are well-structured behavioral guidelines that focus on cost transparency and user assistance, with no evidence of malicious intent, data exfiltration, or harmful prompt injection.

能力评估

✓ Purpose & Capability

The name/description (multi-agent workflow engine with per-run payments) matches the manifest and SKILL.md: it requires a spend token and documents workflow creation, running, scheduling, templates, and agent capabilities. The listed capabilities (search, scraping, sentiment, email, market data, token-analysis, code-audit, etc.) are consistent with a pipeline/orchestration product that delegates work to specialized agents.

✓ Instruction Scope

SKILL.md contains explicit instructions for creating/running/listing/deleting workflows, scheduling, and cost reporting. It instructs the operator to place AIPROX_SPEND_TOKEN in the MCP server config and to use npx/aiprox-workflows-mcp for the MCP adapter. It does not instruct the agent to read unrelated files or arbitrary environment variables, nor to exfiltrate data to unexpected endpoints beyond email/webhook outputs described in the doc.

ℹ Install Mechanism

This is an instruction-only skill (no install spec or code shipped). The docs reference an npm package and an MCP adapter command (npx aiprox-workflows-mcp / npm install aiprox-workflows) but provide no automated installer in the manifest — this is consistent but means installing the adapter is a manual step the user must perform from external registries.

ℹ Credentials

Only AIPROX_SPEND_TOKEN is required, which is proportionate to a billed workflow service. This token is a financial credential (spend token) and should be treated as highly sensitive; the skill doesn't request unrelated tokens or system credentials.

ℹ Persistence & Privilege

always:false (not force-installed). The platform-default ability for the agent to invoke the skill autonomously remains enabled — combined with a spend token this creates a real financial risk (the agent could trigger paid runs). The manifest does not request other privileged system access or modify other skills' config.

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 aiprox-workflows
安装完成后，直接呼叫该 Skill 的名称或使用 /aiprox-workflows 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

v1.0.1

Version 1.0.1 of aiprox-workflows - No file changes detected in this release. - Documentation, features, and usage remain unchanged from the previous version.

v1.0.0

Initial release: Multi-agent AI workflow pipelines for AIProx. - Create, schedule, and manage multi-step workflows that chain AI agents (e.g., search, analytics, email). - Includes commands for workflow creation, running, listing, history, deletion, and template use. - Supports pre-built templates for common use cases (news digest, token scan, competitive intelligence, more). - Pay per execution in sats; no subscriptions, transparent pricing shown for each action. - Requires AIPROX_SPEND_TOKEN for authentication; provides setup instructions. - Documentation includes scheduling, behavior guidelines, and links to dashboard and SDK.

元数据

Slug aiprox-workflows

版本 1.0.1

许可证 MIT-0

累计安装 2

当前安装数 2

历史版本数 2

常见问题