← 返回 Skills 市场
家庭消费意图识别 V4
作者
wuhaichao87
· GitHub ↗
· v4.0.0
323
总下载
0
收藏
1
当前安装
5
版本数
在 OpenClaw 中安装
/install family-expense-intent
功能描述
家庭消费意图识别 V4 - 智能家庭财务管理,支持收入管理、储蓄目标、消费洞察、定期订阅、趋势分析、购物比价。
使用说明 (SKILL.md)
家庭消费意图识别 V4
智能家庭财务管理助手,整合多款优秀技能功能。
功能特性 (V4 新增)
- 🤖 智能识别 - 自动识别消费金额、类别、意图
- 👨👩👧👦 多人管理 - 支持家庭成员独立档案
- 💰 收入管理 - 记录工资、兼职、投资收益
- 🎯 储蓄目标 - 设置目标、追踪进度
- 📊 模式学习 - 从历史记录中学习消费习惯
- ⚠️ 消费洞察 - AI智能分析消费异常
- 🔄 定期订阅 - 管理会员、订阅、房租等周期性支出
- 📈 趋势分析 - 月度/年度消费趋势对比
- 🛒 购物比价 - 淘宝/京东/拼多多比价
- ⚡ 预算管理 - 设置预算,超支提醒
快速开始
1. 初始化
python3 expense_tracker.py add-member "爸爸" "户主"
python3 expense_tracker.py add-member "妈妈" "配偶"
2. 设置预算
python3 expense_tracker.py set-budget "餐饮" 3000
python3 expense_tracker.py set-budget "购物" 2000
3. 记录消费/收入
# 消费(自动NLP)
python3 expense_tracker.py add-conv member_1 "买了件衣服300块"
# 收入
python3 expense_tracker.py add-income member_1 5000 "工资"
# 订阅
python3 expense_tracker.py add-subscription member_1 "Netflix" 30 "娱乐" "月付"
命令参考
消费记录
python3 expense_tracker.py add-conv member_1 "消费描述" [--amount N] [--category X]
python3 expense_tracker.py get-convs [--member X] [--days N] [--category X]
python3 expense_tracker.py stats [--member X] [--days N]
收入管理 (新增)
python3 expense_tracker.py add-income member_1 5000 "工资"
python3 expense_tracker.py add-income member_1 1000 "兼职"
python3 expense_tracker.py get-income [--member X] [--days N]
储蓄目标 (新增)
python3 expense_tracker.py add-goal "旅游基金" 10000 --deadline "2026-12-31"
python3 expense_tracker.py add-goal "紧急备用金" 50000 --priority high
python3 expense_tracker.py contribute-goal "旅游基金" 2000
python3 expense_tracker.py goals
python3 expense_tracker.py goal-progress
定期订阅 (新增)
python3 expense_tracker.py add-subscription member_1 "Netflix" 30 "娱乐" "月付"
python3 expense_tracker.py add-subscription member_1 "房租" 2000 "住房" "月付"
python3 expense_tracker.py subscriptions
python3 expense_tracker.py process-recurring
消费洞察 (新增)
python3 expense_tracker.py insights [--member X]
python3 expense_tracker.py compare-months
趋势分析 (新增)
python3 expense_tracker.py trends 6
python3 expense_tracker.py monthly-report
python3 expense_tracker.py year-report
购物比价 (新增)
python3 expense_tracker.py compare "iPhone 15"
python3 expense_tracker.py compare "吹风机" --source taobao
预算管理
python3 expense_tracker.py set-budget "餐饮" 3000
python3 expense_tracker.py budgets
python3 expense_tracker.py check-budget "餐饮"
python3 expense_tracker.py budget-alerts
储蓄建议 (新增)
python3 expense_tracker.py suggest-savings
消费/收入类别
消费类别
| 类别 | 关键词 |
|---|---|
| 餐饮 | 买菜、做饭、外卖、下馆子 |
| 购物 | 买衣服、网购、电子产品 |
| 交通 | 打车、加油、公交 |
| 教育 | 学费、培训班、买书 |
| 医疗 | 看病、买药、体检 |
| 娱乐 | 电影、游戏、旅游 |
| 通讯 | 话费、网费 |
| 住房 | 房租、水电、物业 |
| 人情 | 红包、礼物、请客 |
收入类别
| 类别 | 说明 |
|---|---|
| 工资 | 主业收入 |
| 兼职 | 副业外快 |
| 投资 | 理财收益 |
| 奖金 | 年终奖等 |
| 其他 | 礼金等 |
数据存储
~/.openclaw/skills-data/family-expense-intent/
├── profiles.json # 家庭成员
├── conversations.json # 消费记录
├── income.json # 收入记录 (V4新增)
├── patterns.json # 消费模式
├── budgets.json # 预算设置
├── goals.json # 储蓄目标 (V4新增)
├── subscriptions.json # 定期订阅 (V4新增)
└── imported_data/ # 导入数据
使用示例
示例1: 记录收入
python3 expense_tracker.py add-income member_1 15000 "工资"
python3 expense_tracker.py add-income member_2 8000 "工资"
示例2: 设置储蓄目标
python3 expense_tracker.py add-goal "买房首付" 300000 --priority high
python3 expense_tracker.py add-goal "旅游" 20000 --deadline "2026-06-01"
python3 expense_tracker.py goal-progress
示例3: 管理订阅
python3 expense_tracker.py add-subscription member_1 "Netflix" 30 "娱乐" "月付"
python3 expense_tracker.py add-subscription member_1 "Apple Music" 10 "娱乐" "月付"
python3 expense_tracker.py subscriptions
示例4: 消费洞察
python3 expense_tracker.py insights
输出示例:
📊 消费洞察:
- ⚠️ 您本月餐饮支出 ¥3500,超出预算 16%
- 💡 您的娱乐支出连续3个月增长,建议关注
- 🏆 本月消费控制良好,超预算类别减少
- 💡 取消Netflix订阅可每月省 ¥30
示例5: 月度对比
python3 expense_tracker.py compare-months
输出示例:
📈 本月 vs 上月:
- 餐饮: +15% ↑
- 购物: -20% ↓
- 交通: 持平
- 总支出: -5% ↓
示例6: 趋势分析
python3 expense_tracker.py trends 6
输出示例:
📊 近6个月趋势:
1月: ¥8,500
2月: ¥9,200 (+8%)
3月: ¥7,800 (-15%)
4月: ¥8,900 (+14%)
5月: ¥8,300 (-7%)
6月: ¥8,500 (+2%)
示例7: 购物比价
python3 expense_tracker.py compare "iPhone 15 256GB"
输出示例:
🛒 iPhone 15 256GB 比价:
┌─────────────┬────────────┬────────────┐
│ 平台 │ 价格 │ 优惠 │
├─────────────┼────────────┼────────────┤
│ 淘宝 │ ¥6,499 │ 无 │
│ 京东 │ ¥6,499 │ 满减200 │
│ 拼多多 │ ¥6,299 │ 百亿补贴 │
└─────────────┴────────────┴────────────┘
推荐: 拼多多 ¥6,299
示例8: 完整月度报告
python3 expense_tracker.py monthly-report
输出:
═══════════════════════════════════════
2026年3月 家庭消费报告
═══════════════════════════════════════
💰 收入: ¥23,000
💸 支出: ¥12,500
💵 结余: ¥10,500 (45%)
📊 支出明细:
餐饮: ¥3,500 (28%) [预算¥3,000, 超16%]
购物: ¥2,000 (16%) [预算¥2,000, 刚好]
教育: ¥2,500 (20%) [预算¥2,500, 刚好]
交通: ¥1,500 (12%) [预算¥1,200, 超25%]
娱乐: ¥1,000 (8%) [预算¥800, 超25%]
其他: ¥2,000 (16%)
📈 趋势:
vs上月: -5%
vs今年均值: +2%
🎯 储蓄目标进度:
买房首付: ¥50,000/¥300,000 (17%)
旅游基金: ¥3,000/¥20,000 (15%)
⚠️ 提醒:
- 餐饮、交通、娱乐超预算
- 建议下月控制餐饮开支
触发方式
消费相关:
- "消费"、"花钱"、"买"、"花了"、"买了"
- "充了"、"交了"、"花了多少"
收入相关:
- "发工资了"、"赚了"、"兼职"、"奖金"
预算相关:
- "预算"、"超支"、"花了多少"
购物相关:
- "想买"、"比价"、"哪个便宜"
目标相关:
- "存钱"、"目标"、"储蓄"
学习自优秀技能
- expense-tracker-pro: 简洁的消费记录
- intelligent-budget-tracker: 收入/目标/订阅/洞察
- taobao: 购物比价
V4 整合升级完成! 🎉
安全使用建议
Before installing or running this skill: 1) Inspect SKILL.md for hidden/unexpected unicode control characters (they can hide instructions or try to manipulate agents/evaluators). 2) Grep the full expense_tracker.py for any network or subprocess usage (requests, urllib, http, socket, subprocess, os.system) — if present, review the remote endpoints and data sent. 3) Verify the "购物比价" feature: either confirm it is a stub/local table or requires external APIs; if it calls external services, ask the author which APIs and why no credentials are declared. 4) Run the tool in an isolated/sandbox environment first (or with a throwaway account) to verify behavior and that it only writes to ~/.openclaw/skills-data/family-expense-intent. 5) Note functional bugs (e.g., get_stats building category stats incorrectly) — treat this as a sign to audit the code before trusting financial reports. 6) Do not provide any system-wide credentials or secret environment variables to this skill until its network behavior and endpoints are audited.
功能分析
Type: OpenClaw Skill
Name: family-expense-intent
Version: 4.0.0
The OpenClaw skill 'family-expense-intent' is a local financial management tool. It stores all data in the designated `~/.openclaw/skills-data/family-expense-intent/` directory. The `SKILL.md` documentation provides clear instructions for using the skill's features without any evidence of prompt injection attempts against the AI agent. The `expense_tracker.py` script performs local file operations and data processing. Crucially, the `compare_product` function, which could potentially involve external network calls for price comparison, is explicitly implemented as a simulation returning hardcoded data, thus posing no current network or external execution risk. There are no indications of data exfiltration, persistence mechanisms, or arbitrary command execution.
能力评估
Purpose & Capability
Name/description and required runtime (python3) align with a local expense-tracking tool. The skill stores data under ~/.openclaw/skills-data/family-expense-intent which is consistent with a local tracker. However, SKILL.md advertises shopping price-compare across Taobao/JD/PDD (networked feature) while the visible portions of expense_tracker.py import only stdlib modules (no requests/urllib) and there are no required credentials declared; this is an inconsistency that could mean the feature is unimplemented, implemented later in the file with hidden network calls, or expects undocumented credentials.
Instruction Scope
Runtime instructions direct the agent/user to run the included python script and operate on local files under the stated data directory — appropriate for this purpose. However, the SKILL.md triggered a pre-scan flag for unicode-control-chars (prompt-injection pattern) which may indicate an attempt to influence evaluators/agents via hidden characters. Also the shipped code contains at least one functional bug (get_stats uses cats.get('cat', ...) which is incorrect), suggesting sloppy maintenance and raising caution about other hidden behaviors.
Install Mechanism
No install spec (instruction-only + included python file). No network-based installer or third-party package installs were specified, so nothing is written beyond the skill's own files and its data directory when run.
Credentials
The skill requests no environment variables or external credentials, which is appropriate for a local tracker. But SKILL.md claims external price-comparison against e-commerce platforms (which commonly require APIs/credentials or scraping); no credentials or network access are declared — this mismatch should be clarified before trusting that feature.
Persistence & Privilege
The skill does not request always:true, does not modify other skills, and stores data only in a subdirectory of the user's home (~/.openclaw/skills-data/family-expense-intent). This is self-contained and proportionate for its stated purpose.
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 family-expense-intent - 安装完成后,直接呼叫该 Skill 的名称或使用
/family-expense-intent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
v4.0.0
family-expense-intent v4.0.0 brings major upgrades for smarter family finance management:
- 全面支持收入管理、储蓄目标、定期订阅、消费洞察和趋势分析。
- 新增购物比价、储蓄建议、月度/年度报告等功能。
- 消费和收入类别扩展,命令行用法大幅丰富。
- 数据结构新增 income.json、goals.json、subscriptions.json,全面支持财务全景。
- 交互和报告更智能,提升家庭财务透明度与洞察力。
v3.1.0
- Major documentation update: SKILL.md rewritten to provide clearer, more concise instructions and usage examples.
- Improved command references with step-by-step usage, parameter explanations, and typical scenarios.
- Streamlined feature list and quick start for easier onboarding.
- Simplified NLP trigger and analysis sections for key supported expressions.
- Privacy, upgrade roadmap, and data structure sections updated for clarity.
v3.0.0
family-expense-intent 3.0.0 introduces major enhancements, including NLP upgrades and new import/OCR features.
- Enhanced NLP recognition for more diverse spending expressions and intent analysis.
- Added support for importing bank/Alipay/WeChat bills and manual entry via expense_tracker.py.
- New OCR functionality: recognize receipt images, auto-extract amount and items, suggest classification.
- Expanded consumption categories, intent types, and budget features.
- Introduced intelligent mode learning for anomalies, frequency, preferences, and personalized suggestions.
- Updated documentation with new usage instructions, output formats, privacy notes, and upgrade roadmap.
v2.0.0
family-expense-intent 2.0.0
- No functional or content changes were made; only metadata (_meta.json) was modified.
- Existing skill description, triggers, and documentation remain unchanged.
v1.0.0
Initial release of family-expense-intent skill:
- Identifies user expense intent, extracts amount, category, urgency, and necessity.
- Supports a wide range of consumption categories (e.g., dining, shopping, housing, education).
- Classifies intent types: necessary, planned, impulsive, enjoyment, or investment.
- Provides structured analysis and practical budget suggestions.
- Handles common real-life family expense scenarios with example responses.
元数据
常见问题
家庭消费意图识别 V4 是什么?
家庭消费意图识别 V4 - 智能家庭财务管理,支持收入管理、储蓄目标、消费洞察、定期订阅、趋势分析、购物比价。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 323 次。
如何安装 家庭消费意图识别 V4?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install family-expense-intent」即可一键安装,无需额外配置。
家庭消费意图识别 V4 是免费的吗?
是的,家庭消费意图识别 V4 完全免费(开源免费),可自由下载、安装和使用。
家庭消费意图识别 V4 支持哪些平台?
家庭消费意图识别 V4 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 家庭消费意图识别 V4?
由 wuhaichao87(@wuhaichao87)开发并维护,当前版本 v4.0.0。
推荐 Skills