← Back to Skills Marketplace
Personal Card Ocr
by
SCNet-sugon
· GitHub ↗
· v1.0.3
· MIT-0
139
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install personal-card-ocr
Description
将图片中的文字、身份证、银行卡、社保卡、户口本、出生医学证明、往来港澳通行证、往来台湾通行证、台湾居民来往大陆通行证、港澳居民来往内地通行证等信息识别并提取出来。本技能应在用户需要 OCR 识别图片中的文字,或识别身份证、银行卡、社保卡、户口本、出生医学证明、往来港澳通行证、往来台湾通行证、台湾居民来往大陆通行证...
Usage Guidance
Before installing, note these practical points:
- Privacy: This skill uploads images (IDs, bank cards, birth certificates, etc.) to the Scnet API (https://api.scnet.cn). If you cannot share such sensitive PII with an external provider, do not use it.
- API key handling: The skill requires SCNET_API_KEY. Prefer setting it as an environment variable rather than placing it in config/.env tracked by a repository; if you use a file, set file permissions (chmod 600) and avoid committing it to source control.
- Provider verification: The skill's homepage in skill.yaml is a placeholder GitHub URL; the source/origin is 'unknown' in the registry metadata. Verify the Scnet service and the repository/maintainer before trusting the API or sending real documents.
- Operational notes: The script respects rate limits and retries on 429. It prints OCR JSON to stdout; the script removes 'confidence' fields from results (harmless but notable).
- Alternatives: If privacy is a priority, consider on-device OCR solutions or an OCR provider with an explicit enterprise privacy / data-retention policy.
If you want more assurance, ask the author for the canonical repository/homepage, their privacy policy, or for an option to run a local/offline OCR model instead.
Capability Analysis
Type: OpenClaw Skill
Name: personal-card-ocr
Version: 1.0.3
The personal-card-ocr skill is a legitimate tool for extracting information from various identification documents using the Sugon-Scnet OCR API. The core logic in scripts/main.py is well-implemented, featuring secure API key handling via configuration files, robust error reporting, and exponential backoff for rate limiting. There is no evidence of malicious behavior, data exfiltration beyond the intended OCR processing, or prompt-injection attacks; the documentation even includes proactive security warnings to prevent users from leaking API keys in chat logs.
Capability Tags
Capability Assessment
Purpose & Capability
Name/description request OCR for personal documents and the skill requires a single SCNET_API_KEY and optional SCNET_API_BASE, and calls an OCR endpoint; the required credential and network access match the stated purpose.
Instruction Scope
SKILL.md and scripts/main.py instruct the agent to read a local image file and POST it to https://api.scnet.cn/api/llm/v1/ocr/recognize. This is within purpose, but the behavior transmits sensitive PII (IDs, bank cards, etc.) to an external service — expected for a cloud OCR integration but high-sensitivity and worth highlighting.
Install Mechanism
No install spec; this is instruction-only plus a small Python script that uses the requests library. No downloads from untrusted URLs or extraction steps are present.
Credentials
Only one required credential (SCNET_API_KEY) and an optional SCNET_API_BASE are declared and actually used. The script reads a config/.env or suggests env var usage; the credential request is proportionate to calling a third-party OCR API.
Persistence & Privilege
always:false and no code modifies other skills or system-wide configs. The skill stores/reads a local config/.env if the user places it there, which is normal and limited in scope.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install personal-card-ocr - After installation, invoke the skill by name or use
/personal-card-ocr - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.3
- Added explicit dependency and environment variable metadata to skill (SKILL.md).
- Now requires Python 3 and the requests library (see dependencies in SKILL.md).
- Improved configuration instructions: only manual .env setup is described; added security warning against sharing API key in chat.
- Added guidance on file permissions for .env.
- Minor clarifications and formatting improvements to documentation for easier usage and troubleshooting.
v1.0.1
Version 1.0.1 of personal-card-ocr
- No user-facing or code changes detected for this release.
- All documentation and implementation remain unchanged compared to the previous version.
- This release may have been made for administrative or versioning purposes.
v0.1.0
- 首次发布个人证件 OCR 技能,支持 9 种主要证件类型识别,包括身份证、银行卡、社保卡、户口本等。
- 提供结构化 JSON 识别结果,涵盖证件关键信息和置信度。
- 支持本地图片文件的 OCR 识别,兼容 jpg、png、pdf 等格式。
- 详细配置指引,需预先申请并配置 Scnet API Token。
- 附带速率限制和错误处理说明,提升调用稳定性与用户体验。
Metadata
Frequently Asked Questions
What is Personal Card Ocr?
将图片中的文字、身份证、银行卡、社保卡、户口本、出生医学证明、往来港澳通行证、往来台湾通行证、台湾居民来往大陆通行证、港澳居民来往内地通行证等信息识别并提取出来。本技能应在用户需要 OCR 识别图片中的文字,或识别身份证、银行卡、社保卡、户口本、出生医学证明、往来港澳通行证、往来台湾通行证、台湾居民来往大陆通行证... It is an AI Agent Skill for Claude Code / OpenClaw, with 139 downloads so far.
How do I install Personal Card Ocr?
Run "/install personal-card-ocr" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Personal Card Ocr free?
Yes, Personal Card Ocr is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Personal Card Ocr support?
Personal Card Ocr is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Personal Card Ocr?
It is built and maintained by SCNet-sugon (@scnet-sugon); the current version is v1.0.3.
More Skills