功能描述

Scans emails and bills to extract billing data, detect refund opportunities, subscription risks, and generate actionable savings recommendations.

使用说明 (SKILL.md)

Skill Definition: claimback_radar

Name: Claimback Radar
Author: schchit

🔑 REQUIRED CREDENTIAL: OPENAI_API_KEY environment variable (or pass api_key directly to ClaimbackRadar).
⚠️ EXTERNAL DATA FLOW: User-provided text content is transmitted to OpenAI's API. Review OpenAI's data retention and privacy policies before use.
🛡️ DATA PRIVACY NOTICE: Do not submit sensitive financial or identity data unless you explicitly accept this external data flow.

Description

Scans user emails and bills to discover hidden refunds, subscription traps, and savings opportunities.

Functions

Function 1: extract

Extracts structured subscription / billing data from unstructured text.

Input: Raw email or bill text Output: confirmation_card (JSON)

Function 2: detect_and_recommend

Detects risks and generates actionable receipts.

Input: confirmation_card + user context Output: action_receipts + risk_flags

Invocation Schema

See schema/input.json

Output Schema

See schema/output.json

Example

See examples/netflix_email.md

Security Notes

The CLI entrypoint (main.py) explicitly loads .env if present and warns the user.
Do not commit .env files containing real API keys to version control.
For production use, prefer explicit api_key injection over implicit environment loading.
Ensure no unrelated secrets are stored in .env in directories where you run this tool.

安全使用建议

This skill is coherent and appears to do what it says: it sends the email/bill text you provide to OpenAI to extract billing/subscription information and generate recommendations. Before installing or using it, consider: 1) Privacy — any text you feed (emails, bills) will be transmitted to OpenAI; do not include highly sensitive PII or financial account details unless you accept OpenAI's data handling. 2) .env behavior — main.py will load a local .env if present (it warns you), so ensure that file does not contain unrelated secrets you don't want read at runtime. 3) Supply OPENAI_API_KEY securely (prefer explicit api_key injection over leaving keys in .env). 4) If you need stricter guarantees, review the prompts (prompts/system.txt) and run the tests locally to confirm behavior. Overall: functionality is proportionate, main risk is expected external data exposure — evaluate whether that privacy tradeoff is acceptable for your use case.

功能分析

Type: OpenClaw Skill Name: claimback-radar Version: 1.0.5 The claimback-radar skill is an AI-powered tool designed to extract billing and subscription data from text using the OpenAI API. The implementation is transparent, featuring clear documentation in SKILL.md and README.md, explicit privacy warnings regarding external data flow, and a manifest (clawhub.json) that correctly identifies its network requirements. The code (main.py, runner.py) follows its stated purpose without any signs of malicious execution, obfuscation, or unauthorized data exfiltration.

能力标签

cryptocan-make-purchasesrequires-sensitive-credentials

能力评估

✓ Purpose & Capability

Name/description (scan emails/bills, detect refunds/subscriptions) maps to the code and declared requirements. The package depends only on the OpenAI SDK, python-dotenv and pydantic which are reasonable for the described LLM-driven extraction tasks. Required credential OPENAI_API_KEY is expected and used by src/runner.py.

ℹ Instruction Scope

Runtime instructions and code explicitly transmit user-provided email/bill text to OpenAI for processing (this is documented and warned about in SKILL.md and README). The code only reads the provided input JSON and the local system prompt file; it does not access unrelated system files or additional credentials. The notable scope issue is the privacy impact of sending potentially sensitive financial/identity data to an external API — the skill does warn users not to submit sensitive info.

✓ Install Mechanism

No custom or remote arbitrary downloads. The declared install path is pip installing requirements.txt (openai, python-dotenv, pydantic). This is a standard package install model with traceable dependencies.

✓ Credentials

Only OPENAI_API_KEY is required (declared in clawhub.json and SKILL.md) which is proportionate to the LLM usage. The only other runtime behavior is optional loading of a local .env file (main.py uses python-dotenv) — this is warned in multiple places but still a privacy footgun if users have unrelated secrets in .env.

✓ Persistence & Privilege

No 'always' privilege requested. The skill does not persist or modify other skills or system-wide configs. It reads input files and system prompt, and makes outbound calls only to OpenAI (as documented).

版本历史

v1.0.5

claimback-radar 1.0.5 - Added project configuration files: `clawhub.json` and `pyproject.toml`. - Updated data privacy notice to include identity data, not just financial data. - Expanded security notes to warn against storing unrelated secrets in `.env`.

v1.0.4

claimback-radar 1.0.4 - Updated SKILL.md to add a clearer data privacy notice and visual highlights for key security information. - Added _meta.json metadata file. - Removed pyproject.toml.

v1.0.3

# Changelog All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [1.0.3] - 2026-04-23 ### Fixed - Unified package version to `1.0.3` across `pyproject.toml` and registry metadata - Added prominent `OPENAI_API_KEY` credential declaration to `README.md` and `SKILL.md` first lines - Removed unused `pydantic` dependency from `requirements.txt` and `pyproject.toml` - Clarified external data flow warning (OpenAI API) in registry metadata ## [0.1.0] - 2026-04-22 ### Added - **Function 1: Structured Extraction** — Scans raw email/bill text and extracts service name, provider, amount, billing cycle, next charge date, refund deadline, warranty expiry, and cancellation method - **Function 2: Risk Detection & Action Generation** — Detects risks (upcoming charges, refund windows closing, unused services, price hikes, warranty expiring) and generates prioritized action receipts with deadlines and estimated savings - **Schema definitions** — `schema/input.json` and `schema/output.json` for strict validation of Skill invocation and response - **System prompt** — `prompts/system.txt` containing the core extraction and detection logic for LLM execution - **Reference implementation** — `src/runner.py` with `ClaimbackRadar` class, runnable via `main.py` CLI or imported as a module - **Examples** — Netflix subscription and Amazon Prime renewal cases with expected input/output - **Unit tests** — `tests/test_runner.py` covering prompt loading, valid output structure, and error handling - **Paper reference** — Links to Zenodo preprint (DOI: 10.5281/zenodo.19678205) and MDPI *Information* submission [1.0.3]: https://github.com/schchit/claimback-radar/releases/tag/v1.0.3 [0.1.0]: https://github.com/schchit/claimback-radar/releases/tag/v0.1.0

v1.0.2

# Changelog All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [0.1.1] - 2026-04-22 ### Fixed - Added prominent `OPENAI_API_KEY` credential declaration to `README.md` and `SKILL.md` first lines - Removed unused `pydantic` dependency from `requirements.txt` - Clarified external data flow warning (OpenAI API) in registry metadata ## [0.1.0] - 2026-04-22 ### Added - **Function 1: Structured Extraction** — Scans raw email/bill text and extracts service name, provider, amount, billing cycle, next charge date, refund deadline, warranty expiry, and cancellation method - **Function 2: Risk Detection & Action Generation** — Detects risks (upcoming charges, refund windows closing, unused services, price hikes, warranty expiring) and generates prioritized action receipts with deadlines and estimated savings - **Schema definitions** — `schema/input.json` and `schema/output.json` for strict validation of Skill invocation and response - **System prompt** — `prompts/system.txt` containing the core extraction and detection logic for LLM execution - **Reference implementation** — `src/runner.py` with `ClaimbackRadar` class, runnable via `main.py` CLI or imported as a module - **Examples** — Netflix subscription and Amazon Prime renewal cases with expected input/output - **Unit tests** — `tests/test_runner.py` covering prompt loading, valid output structure, and error handling - **Paper reference** — Links to Zenodo preprint (DOI: 10.5281/zenodo.19678205) and MDPI *Information* submission [0.1.1]: https://github.com/schchit/claimback-radar/releases/tag/v0.1.1 [0.1.0]: https://github.com/schchit/claimback-radar/releases/tag/v0.1.0

v1.0.1

claimback-radar v1.0.1 - Added requirements for `OPENAI_API_KEY` and clarified API key usage instructions. - Added explicit data privacy notice about sending user content to OpenAI's API. - Included security guidance regarding `.env` usage, warning against committing secrets and preferring explicit API key passing. - No changes to code functionality.

v1.0.0

# Changelog All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). ## [0.1.0] - 2026-04-22 ### Added - **Function 1: Structured Extraction** — Scans raw email/bill text and extracts service name, provider, amount, billing cycle, next charge date, refund deadline, warranty expiry, and cancellation method. - **Function 2: Risk Detection & Action Generation** — Detects risks (upcoming charges, refund windows closing, unused services, price hikes, warranty expiring) and generates prioritized action receipts with deadlines and estimated savings. - **Schema definitions** — `schema/input.json` and `schema/output.json` for strict validation of Skill invocation and response. - **System prompt** — `prompts/system.txt` containing the core extraction and detection logic for LLM execution. - **Reference implementation** — `src/runner.py` with `ClaimbackRadar` class, runnable via `main.py` CLI or imported as a module. - **Examples** — Netflix subscription and Amazon Prime renewal cases with expected input/output. - **Unit tests** — `tests/test_runner.py` covering prompt loading, valid output structure, and error handling. - **Paper reference** — Links to Zenodo preprint (DOI: 10.5281/zenodo.19678205) and MDPI *Information* submission. [0.1.0]: https://github.com/schchit/claimback-radar/releases/tag/v0.1.0

元数据

Slug claimback-radar

版本 1.0.5

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 6

常见问题

Claimback Radar 是什么？

Scans emails and bills to extract billing data, detect refund opportunities, subscription risks, and generate actionable savings recommendations. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 123 次。

如何安装 Claimback Radar？

在 OpenClaw 或 Claude Code 对话框中运行命令「/install claimback-radar」即可一键安装，无需额外配置。

Claimback Radar 是免费的吗？

是的，Claimback Radar 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。

Claimback Radar 支持哪些平台？

Claimback Radar 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 Claimback Radar？

由 JEP (Judgment Event Protocol)（@schchit）开发并维护，当前版本 v1.0.5。

Claimback Radar