← 返回 Skills 市场
rishabhdugar

Document Similarity

作者 Rishabh Dugar · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ 安全检测通过
84
总下载
1
收藏
1
当前安装
2
版本数
在 OpenClaw 中安装
/install document-similarity
功能描述
Compare two images or PDFs for visual similarity via the PDFAPIHub cloud API. Documents are uploaded to pdfapihub.com for comparison using feature matching,...
使用说明 (SKILL.md)

Document Similarity

What It Does

Compares two documents (images or PDFs) for visual similarity via the PDFAPIHub hosted API. Both documents are uploaded to PDFAPIHub servers where comparison is performed, and a similarity score (0–1) with confidence level is returned.

When to Use

  • Check if two documents are visually similar
  • Detect duplicates or near-duplicates
  • Compare image variations

Comparison Methods

Method Description
auto Automatically selects best method (default)
feature_match OpenCV feature matching
ssim Structural Similarity Index
phash Perceptual hashing

Supported Combinations

  • image + image
  • pdf + pdf
  • image + pdf

Required Inputs

Two files via one of:

  • url1 + url2 — public URLs
  • image1_base64 + image2_base64 — base64-encoded files
  • Multipart upload with file1 and file2

Authentication

This skill calls the PDFAPIHub hosted API at https://pdfapihub.com/api. Both documents are uploaded to PDFAPIHub servers for comparison.

Send your API key in the CLIENT-API-KEY header.

Get your free API key at https://pdfapihub.com. Full API documentation is available at https://pdfapihub.com/docs.

Privacy note: Both documents you compare are uploaded to PDFAPIHub's cloud service for processing. Do not send confidential documents unless you trust the service. Files are auto-deleted after 30 days.

Use Cases

  • Duplicate Detection — Identify duplicate or near-duplicate documents in a repository
  • Brand Consistency — Compare generated documents against approved templates for visual consistency
  • QA Testing — Compare rendered PDFs/images before and after code changes for regressions
  • Fraud Detection — Compare submitted documents against known genuine samples
  • Document Versioning — Quantify visual differences between document revisions
  • Container Inspection — Compare shipping container photos for damage assessment

Example Usage

curl -X POST https://pdfapihub.com/api/v1/document/similarity \
  -H "CLIENT-API-KEY: your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "url1": "https://pdfapihub.com/sample-document-similarity-1.jpg",
    "url2": "https://pdfapihub.com/sample-document-similarity-2.jpg",
    "method": "auto"
  }'
安全使用建议
This skill appears internally consistent and simply proxies two documents to PDFAPIHub for comparison. Before using it: (1) Do not upload confidential or regulated documents unless you trust pdfapihub.com and have reviewed their privacy/retention policies (SKILL.md says files are auto-deleted after 30 days). (2) Provide an API key in the CLIENT-API-KEY header when invoking; note the registry metadata omitted a declared primary credential — confirm how your deployment will supply the key (per-request header vs. environment variable). (3) Test with non-sensitive sample files first to verify behavior. (4) If you need offline/local comparisons or do not want third-party uploads, this skill is not appropriate.
能力标签
requires-sensitive-credentials
能力评估
Purpose & Capability
The skill's name, description, SKILL.md and example.json consistently describe uploading two documents to https://pdfapihub.com/api for similarity comparison. One small inconsistency: registry metadata lists no primary credential or required env vars, while skill.json and SKILL.md indicate an API key (CLIENT-API-KEY) is required for requests. This is likely an administrative/metadata omission rather than a functional mismatch.
Instruction Scope
The runtime instructions only describe sending the two files (via URLs, base64 bodies, or multipart) to the PDFAPIHub API and receiving a similarity score. There are no instructions to read local system files, secrets, or unrelated config, nor to transmit data to any endpoint other than pdfapihub.com.
Install Mechanism
This is an instruction-only skill with no install spec and no code files to write to disk, so there is no install-time risk.
Credentials
The service requires an API key sent in the CLIENT-API-KEY header, which is appropriate for this cloud API. However, the skill registry metadata did not list a primary credential or required env var while skill.json marks auth as required — an inconsistency that should be clarified (the skill itself does not attempt to access other unrelated credentials).
Persistence & Privilege
The skill does not request persistent presence (always: false), does not modify other skills or system settings, and has no install-time hooks. Autonomous invocation is enabled by default but is not combined with other concerning privileges.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install document-similarity
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /document-similarity 触发
  4. 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
Clarify hosted API (documents uploaded to PDFAPIHub for comparison). Add privacy notice. Declare CLIENT-API-KEY credential in description.
v1.0.0
Compare two images or PDFs and return a similarity score (0-1) with confidence level. Methods: feature matching (OpenCV), SSIM, perceptual hashing. Supports image+image, PDF+PDF, and image+PDF combinations.
元数据
Slug document-similarity
版本 1.0.1
许可证 MIT-0
累计安装 1
当前安装数 1
历史版本数 2
常见问题

Document Similarity 是什么?

Compare two images or PDFs for visual similarity via the PDFAPIHub cloud API. Documents are uploaded to pdfapihub.com for comparison using feature matching,... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 84 次。

如何安装 Document Similarity?

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

Document Similarity 是免费的吗?

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

Document Similarity 支持哪些平台?

Document Similarity 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 Document Similarity?

由 Rishabh Dugar(@rishabhdugar)开发并维护,当前版本 v1.0.1。

💬 留言讨论