rent

通过 RentAHuman.ai 将任务委托给真人 - 搜索技能、发布赏金、管理对话以及运行人工智能评分的机会扫描。

# Rent-A-Human Bounty Hunter

Scans RentAHuman.ai bounties via MCP + API. Uses Grok AI to filter spam,
score opportunities by location, skills, and ease of completion, and sends
top results to Telegram.

🔍 Scan Bounties - Find and score job opportunities based on your location and skills

🧑‍💼 Browse Humans - View available humans for hire with skills and rates

💼 View Bounties - See open jobs and opportunities

📝 Post Jobs - Create new bounties directly from Telegram

🎯 Smart Scanning - AI-scored opportunity recommendations

🔒 Private - Bot only responds to your user ID

🔗 MCP + API Integration - Direct connection to official RentAHuman.ai MCP

Rent-A-Human MCP Server Integration: Full agentic access to the RentAHuman.ai platform

CLI Interface: Fully customizable /rent agent mode

AI-Powered Scoring: Uses Grok-4-1-fast-reasoning to score job opportunities on a scale of 0-100

Smart Caching: Caches results for 12 hours to avoid redundant API calls

Telegram Integration: Sends top opportunities directly to your Telegram

Drag & Drop Skills: Compatible with Claude Code and OpenClaw — drop the .claude/skills/ folder into any project

Multiple Scan Modes:

Cached mode (fast, uses existing scores)
Force fresh scoring (bypass cache)
List all open jobs
List available humans for hire

## Requirements
- `XAI_API_KEY` — from x.ai for Grok scoring
- `RENTAHUMAN_API_KEY` — from rentahuman.ai/dashboard
- `TELEGRAM_BOT_TOKEN` (optional) — for Telegram notifications
- `TELEGRAM_CHAT_ID` (optional) — for Telegram notifications

# Rent-A-Human Skill

Auto-activate when the user wants to hire a person, post a job, or delegate a task that needs a human.

## Quick Commands

```
/rent              — Browse + print command menu
/rent scan         — AI-scored bounties (Grok, 12hr cache)
/rent scan force   — Bypass cache, fresh scoring now
/rent scan new     — Only unseen bounties
/rent post <desc>  — Post a new bounty (first sentence = title)
/rent saved        — View saved bounties
/rent skills       — List available human skills
/rent status       — Connection check
```

## Hire Flow

**Direct:** `/rent search <skill>` → `/rent human <id>` → `/rent talk <id> : msg` → negotiate → book  
**Bounty:** `/rent post desc` → humans apply → `/rent applications <id>` → `/rent accept <app_id>`

## Bounty Scanner

Script: `python3 .claude/skills/rent/scripts/bounty_hunter.py`  
Cache: `logs/bounties_cache.json` (12hr TTL, Grok-3-mini-fast scoring 0–100)  
After running: read cache file and display scored results to user.

## Rate Limits

Bounties: 5/day | Conversations: 50/day | Messages: 30/hr | API keys: 3 max

## CLI Alternative

```bash
python hire_team.py "Task description"
python hire_team.py talk <human_id> "Message"
python hire_team.py bounty "Title" --description "..." --price 100 --hours 2
```

**Note:** `hire_team.py` is an optional external CLI. Primary interface is MCP via `/rent` commands.

Payment via Stripe Connect escrow on RentAHuman.ai.

→ Full command tree + MCP tool map: [`.claude/commands/rent.md`](.claude/commands/rent.md)

---

## 中文说明

# Rent-A-Human 赏金猎手

通过 MCP + API 扫描 RentAHuman.ai 上的赏金任务。使用 Grok AI 过滤垃圾信息，
按地点、技能和完成难易程度对机会进行评分，并将最优结果发送到 Telegram。

🔍 扫描赏金 - 根据你的地点和技能查找并评分工作机会

🧑‍💼 浏览人选 - 查看可雇佣的人选及其技能与价格

💼 查看赏金 - 查看开放的工作和机会

📝 发布任务 - 直接从 Telegram 创建新赏金

🎯 智能扫描 - AI 评分的机会推荐

🔒 私密 - 机器人仅响应你的用户 ID

🔗 MCP + API 集成 - 直连官方 RentAHuman.ai MCP

Rent-A-Human MCP 服务器集成：对 RentAHuman.ai 平台的完整代理式访问

CLI 界面：完全可自定义的 /rent 代理模式

AI 驱动评分：使用 Grok-4-1-fast-reasoning 在 0-100 区间内为工作机会评分

智能缓存：缓存结果 12 小时，避免重复的 API 调用

Telegram 集成：将最优机会直接发送到你的 Telegram

拖放式技能：兼容 Claude Code 和 OpenClaw — 将 .claude/skills/ 文件夹拖入任意项目即可

多种扫描模式：

缓存模式（快速，使用已有评分）
强制重新评分（绕过缓存）
列出所有开放任务
列出可雇佣的人选

## 要求
- `XAI_API_KEY` — 来自 x.ai，用于 Grok 评分
- `RENTAHUMAN_API_KEY` — 来自 rentahuman.ai/dashboard
- `TELEGRAM_BOT_TOKEN`（可选）— 用于 Telegram 通知
- `TELEGRAM_CHAT_ID`（可选）— 用于 Telegram 通知

# Rent-A-Human 技能

当用户想要雇佣某人、发布任务或委派需要真人完成的任务时自动激活。

## 快捷命令

```
/rent              — 浏览 + 打印命令菜单
/rent scan         — AI 评分的赏金（Grok，12 小时缓存）
/rent scan force   — 绕过缓存，立即重新评分
/rent scan new     — 仅未查看过的赏金
/rent post <desc>  — 发布新赏金（首句作为标题）
/rent saved        — 查看已保存的赏金
/rent skills       — 列出可用的人选技能
/rent status       — 连接检查
```

## 雇佣流程

**直接：** `/rent search <skill>` → `/rent human <id>` → `/rent talk <id> : msg` → 协商 → 预订  
**赏金：** `/rent post desc` → 人选申请 → `/rent applications <id>` → `/rent accept <app_id>`

## 赏金扫描器

脚本：`python3 .claude/skills/rent/scripts/bounty_hunter.py`  
缓存：`logs/bounties_cache.json`（12 小时 TTL，Grok-3-mini-fast 评分 0–100）  
运行后：读取缓存文件并向用户展示评分结果。

## 速率限制

赏金：5/天 | 对话：50/天 | 消息：30/小时 | API 密钥：最多 3 个

## CLI 替代方案

```bash
python hire_team.py "Task description"
python hire_team.py talk <human_id> "Message"
python hire_team.py bounty "Title" --description "..." --price 100 --hours 2
```

**注意：** `hire_team.py` 是可选的外部 CLI。主要界面是通过 `/rent` 命令的 MCP。

通过 RentAHuman.ai 上的 Stripe Connect 托管进行付款。

→ 完整命令树 + MCP 工具映射：[`.claude/commands/rent.md`](.claude/commands/rent.md)