{"success":true,"data":[{"skill_name":"pdf","skill_description":"处理 PDF 文件：提取文本、创建 PDF、合并文档等；适用于用户要读/生成/合并 PDF 或进行 PDF 相关处理。","relevance_score":57.2389,"author":"anymouschina","stars":301,"skill_url":"https://github.com/anymouschina/TapCanvas/blob/418948a19d28c08e236f837d2918b2a4ea64bbb4/apps/agents-cli/skills/pdf","category":"Development","evaluation":{"safety":{"level":"Good","reason":"The skill deals with benign PDF file operations (reading, creating, merging, splitting) with no destructive actions, security bypasses, or risky automation. No safety concerns identified."},"completeness":{"level":"Good","reason":"The SKILL.md covers core PDF operations with multiple methods (pdftotext, PyMuPDF, ReportLab, pandoc), provides concrete code examples, installation commands, and best practices. While password-protected PDFs and error handling for corrupted files aren't mentioned, the essential workflows are well-documented."},"executability":{"level":"Good","reason":"The skill provides concrete, actionable instructions with specific commands (pdftotext, pandoc) and runnable Python code snippets using well-established libraries (PyMuPDF, ReportLab). An agent can follow these steps to perform PDF operations without ambiguity."},"cost_awareness":{"level":"Good","reason":"The skill explicitly recommends processing large PDFs page-by-page to avoid memory issues and mentions using appropriate tools for specific tasks. The operations are inherently lightweight file processing without external API calls or heavy compute requirements."},"maintainability":{"level":"Good","reason":"The skill has narrow, modular scope focused on PDF operations with clear separation between reading, creating, merging, and splitting. Configuration points are mentioned (e.g., custom styling with pandoc, page-by-page processing for large PDFs), making it easy to adapt or extend."}},"repo_name":"anymouschina/TapCanvas","skill_dir":"anymouschina/TapCanvas/apps/agents-cli/skills/pdf"},{"skill_name":"Pdf","skill_description":"USE WHEN pdf, PDF file.","relevance_score":56.0651,"author":"edheltzel","stars":2,"skill_url":"https://github.com/edheltzel/atlas/blob/96b9b06f12c09b0a2c08a81336407ece1905f521/claudecode/.claude/skills/Documents/Pdf","category":"Productivity","evaluation":{"safety":{"level":"Good","reason":"The skill covers benign PDF operations (merge, split, OCR, form filling, watermarking) and does not instruct destructive or credential-exfiltration actions; examples (pypdf, pdfplumber, pytesseract) focus on file processing and encryption rather than unsafe behavior."},"completeness":{"level":"Good","reason":"SKILL.md documents multiple concrete workflows with code and CLI examples for creation, merge/split, text/table extraction, OCR, and manipulation and includes a detailed forms.md describing required JSON formats and step-by-step procedures, providing clear inputs, tools, and outputs for common PDF tasks."},"executability":{"level":"Average","reason":"Most workflows include runnable code and CLI commands (e.g., pypdf, pdfplumber, pdftotext), but forms.md instructs running helper scripts (e.g., scripts/extract_form_field_info.py, scripts/convert_pdf_to_images.py) while the provided SCRIPTS/ listing shows no scripts, creating a practical gap for some tasks."},"cost_awareness":{"level":"Average","reason":"The documentation is silent about limits, batching, or large-scale processing controls (e.g., OCR converts all pages via pdf2image without advising sampling or rate limits), so while typical single-file use is low-cost, large/batch workloads have no cost-management guidance."},"maintainability":{"level":"Good","reason":"The skill is modularly organized into distinct workflows (Creation, Merge/Split, Text/Table extraction, OCR, Manipulation) with clear tool choices (reportlab, pypdf, pdfplumber, pytesseract) and explicit code snippets, making it straightforward to adapt or extend for other PDF use cases."},"script_execution":[]},"repo_name":"edheltzel/atlas","skill_dir":"edheltzel/atlas/claudecode/.claude/skills/Documents/Pdf"},{"skill_name":"pdf","skill_description":"Process PDF files — read, create, merge, split, rotate, watermark, extract text/tables/images, fill forms, encrypt/decrypt, and OCR scanned PDFs. Use this skill whenever a .pdf file is involved as input or output, or when the user mentions PDF processing. Triggers on 'PDF', '.pdf', 'merge PDF', 'split PDF', 'extract text', 'fill form', 'create PDF', 'PDF 생성', 'PDF 합치기', 'PDF 나누기', 'PDF 양식', '텍스트 추출'.","relevance_score":55.9056,"author":"berrzebb","stars":7,"skill_url":"https://github.com/berrzebb/quorum/blob/29a1f4b4f44c5b0110468ee068ad5a3bd483747d/skills/pdf","category":"Development","evaluation":{"safety":{"level":"Good","reason":"The skill operates on benign PDF processing tasks and does not include destructive or exfiltration behaviors; SKILL.md and references focus on read/create/merge/split/ocr and form filling without instructing unsafe actions."},"completeness":{"level":"Average","reason":"SKILL.md and references describe detailed workflows, required libraries, and explicit form-filling steps (references/forms.md), but several bundled scripts in scripts/ are truncated or incomplete (e.g., fill_pdf_form_with_annotations.py, extract_form_structure.py), leaving gaps in the provided implementation."},"executability":{"level":"Average","reason":"The skill gives concrete commands, scripts, and library installation lines, however multiple shipped scripts are syntactically truncated or cut off (create_validation_image.py, extract_form_structure.py, fill_pdf_form_with_annotations.py) so out-of-the-box execution would fail without fixing those files."},"cost_awareness":{"level":"Average","reason":"The documentation lists resource-intensive tools for OCR (pdf2image, pytesseract) and rendering but does not provide explicit guidance on batching, page or file-size limits, or cost-control strategies for large or many PDFs."},"maintainability":{"level":"Average","reason":"The skill is modular with clear scripts and reference docs (references/reference.md, references/forms.md), but the presence of incomplete scripts and some hard-coded patterns means additional work is required to adapt and reliably maintain the implementation."}},"repo_name":"berrzebb/quorum","skill_dir":"berrzebb/quorum/skills/pdf"},{"skill_name":"Pdf","skill_description":"PDF processing. USE WHEN pdf, PDF file. SkillSearch('pdf') for docs.","relevance_score":55.6584,"author":"danielmiessler","stars":9762,"skill_url":"https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/903d2bd57dea9b33cff38917c95b7317a610ae67/Releases/v2.5/.claude/skills/Documents/Pdf","category":"Productivity","evaluation":{"safety":{"level":"Average","reason":"The skill covers benign PDF operations but includes potentially sensitive actions (e.g., qpdf example showing password removal: \"qpdf --password=mypassword --decrypt encrypted.pdf decrypted.pdf\") and provides no explicit safeguards, confirmation steps, or guidance about handling sensitive content."},"completeness":{"level":"Good","reason":"SKILL.md provides clear workflows, triggers, tools, and concrete code/CLI examples for creation, merge/split, text/table extraction, OCR, form filling, and manipulation, and points to forms.md/reference.md for detailed procedures, covering essential steps and expected outputs."},"executability":{"level":"Good","reason":"The document includes concrete, ready-to-run code snippets and CLI commands (e.g., pypdf merge/split examples, pdfplumber table extraction, pdftotext/qpdf commands) so an agent with typical Python/CLI tooling can follow them, although some referenced helper scripts (forms.md scripts) are not included in the package."},"cost_awareness":{"level":"Average","reason":"The guide does not discuss batching, page/size limits, or performance/cost trade-offs for expensive operations like OCR or large-scale processing (OCR examples exist but no limits or recommendations are provided)."},"maintainability":{"level":"Good","reason":"Workflows are modular and tied to common libraries (pypdf, pdfplumber, reportlab, pytesseract) with example snippets that show where to change inputs/parameters, making it straightforward to adapt or compose into other pipelines."},"script_execution":[]},"repo_name":"danielmiessler/Personal_AI_Infrastructure","skill_dir":"danielmiessler/Personal_AI_Infrastructure/Releases/v2.5/.claude/skills/Documents/Pdf"},{"skill_name":"pdf","skill_description":"Create, merge, split, and extract content from PDF files. Activates for \"PDF\", \"merge PDF\", \"split PDF\", \"extract text\", \"create PDF\", \"PDF form\".","relevance_score":55.4033,"author":"abdullah1854","stars":12,"skill_url":"https://github.com/abdullah1854/MCPGateway/blob/9ce24d139068e08e49ba82364a913ec08fe30ee4/.agents/skills/pdf","category":"Productivity","evaluation":{"safety":{"level":"Average","reason":"The SKILL.md covers benign PDF operations but also shows potentially sensitive actions like qpdf --decrypt and form filling without any usage safeguards or data-privacy guidance, which raises moderate misuse/privacy risk."},"completeness":{"level":"Good","reason":"The document provides concrete workflows, code snippets for extract/merge/split/create/rotate/watermark/password-protect/OCR/form filling and lists dependencies and CLI alternatives, covering the essential steps and prerequisites."},"executability":{"level":"Good","reason":"Multiple concrete Python and CLI examples (pdfplumber, pypdf, reportlab, qpdf, pdftotext, pytesseract) and dependency install commands are provided, so an agent with standard tools can realistically run the described workflows."},"cost_awareness":{"level":"Average","reason":"The SKILL.md lists OCR and image conversion workflows (pdf2image, pytesseract) that can be compute-heavy but does not mention batching, page limits, or other cost controls to limit expensive operations."},"maintainability":{"level":"Good","reason":"The skill is modular with clear library recommendations and separate examples for each operation (merge, split, extract, create, OCR, form filling), making it straightforward to adapt or extend for new use cases."},"script_execution":[]},"repo_name":"abdullah1854/MCPGateway","skill_dir":"abdullah1854/MCPGateway/.agents/skills/pdf"},{"skill_name":"pdf","skill_description":"このスキルは、ユーザーが「PDFを作成」「PDFからテキストを抽出」「PDFを結合」「PDFフォームに記入」「PDFを分割」「PDFからテーブルを抽出」「PDF OCR」「PDFを暗号化」「PDFメタデータを編集」などを要求した際に使用する。\n","relevance_score":55.2851,"author":"tadokoro-ryusuke","stars":0,"skill_url":"https://github.com/tadokoro-ryusuke/cc-plugins/blob/8a6a48c952e0f3fea65947fc5090db777d632010/ms-office-suite/skills/pdf","category":"Productivity","evaluation":{"safety":{"level":"Good","reason":"All described operations are benign PDF manipulations (text extraction, merging, OCR, form filling, encryption) and the SKILL.md and scripts do not include destructive or network-exfiltration behaviors; example tools are standard (pypdf, qpdf, pdftotext) and no unsafe commands or credential-bypassing steps are suggested."},"completeness":{"level":"Average","reason":"The SKILL.md covers many PDF tasks with commands, libraries, and a detailed form-filling workflow (references/forms.md) and dependency list, but several script snippets in SCRIPTS are truncated/incomplete in the provided content and there is limited discussion of error handling and edge cases for large or malformed PDFs."},"executability":{"level":"Average","reason":"Concrete commands and Python examples (pdftotext, pdfplumber, reportlab, qpdf, scripts for form workflows) make the workflow actionable and several scripts compiled successfully, but multiple script files shown in SCRIPTS are truncated/incomplete in the provided snippets which creates uncertainty about fully runnable end-to-end execution."},"cost_awareness":{"level":"Average","reason":"The SKILL.md lists heavier operations (OCR with pdf2image+pytesseract, rendering with pypdfium2) but does not specify batching, page/size limits, or guidelines to limit expensive processing, so it lacks explicit cost-control advice despite potentially costly workloads."},"maintainability":{"level":"Average","reason":"The skill is modular (separate scripts for checking fillable fields, extracting form info, adding annotations) and uses standard libraries, which aids reuse, but some scripts are tightly coupled to repo-specific paths (e.g., ${CLAUDE_PLUGIN_ROOT}/skills/pdf/scripts) and the provided snippets are partially incomplete, increasing the effort to adapt."},"script_execution":[{"note":"no usage examples found; py_compile succeeded","path":"scripts/check_bounding_boxes.py","error":null,"status":"compiled_only","command":"python -m py_compile scripts/check_bounding_boxes.py","exit_code":0,"duration_sec":0.055},{"note":"no usage examples found; py_compile succeeded","path":"scripts/check_fillable_fields.py","error":null,"status":"compiled_only","command":"python -m py_compile scripts/check_fillable_fields.py","exit_code":0,"duration_sec":0.056},{"note":"no usage examples found; py_compile succeeded","path":"scripts/convert_pdf_to_images.py","error":null,"status":"compiled_only","command":"python -m py_compile scripts/convert_pdf_to_images.py","exit_code":0,"duration_sec":0.036},{"note":"no usage examples found; py_compile succeeded","path":"scripts/create_validation_image.py","error":null,"status":"compiled_only","command":"python -m py_compile scripts/create_validation_image.py","exit_code":0,"duration_sec":0.036},{"note":"no usage examples found; py_compile succeeded","path":"scripts/extract_form_field_info.py","error":null,"status":"compiled_only","command":"python -m py_compile scripts/extract_form_field_info.py","exit_code":0,"duration_sec":0.035}]},"repo_name":"tadokoro-ryusuke/cc-plugins","skill_dir":"tadokoro-ryusuke/cc-plugins/ms-office-suite/skills/pdf"},{"skill_name":"pdf","skill_description":"PDF全功能处理：读取/提取、合并/拆分、水印、表单填写、加密、OCR。关键词：.pdf、PDF、合并PDF、拆分PDF、PDF表单、水印、生成PDF","relevance_score":55.1818,"author":"senweaver","stars":15,"skill_url":"https://github.com/senweaver/senweaver-ide/blob/509d73a2bd1243248e5c149be7d674d92333a92a/.senweaver/skill/pdf","category":"Productivity","evaluation":{"safety":{"level":"Good","reason":"The guide operates on local PDF files with benign actions (merge, split, watermark, OCR) and does not include destructive or credential-exfiltration steps; examples use libraries/CLI tools like pypdf, pdfplumber, qpdf, and pdftotext on local paths."},"completeness":{"level":"Good","reason":"The SKILL.md covers a wide range of PDF tasks with concrete code snippets and prerequisites (pypdf, pdfplumber, reportlab, OCR tooling), includes form-filling guidance in FORMS.md, and notes issues such as XFA forms and ReportLab font caveats."},"executability":{"level":"Good","reason":"There are concrete, runnable examples for Python (pypdf/pdfplumber/reportlab) and CLI tools (pdftotext, qpdf) and installation commands, enabling execution; although scripts/pdf_utils.py is present, its split_pdf function is truncated, most documentation snippets are complete and actionable."},"cost_awareness":{"level":"Average","reason":"The guide lists required heavy-tooling for OCR and conversions (pytesseract, pdf2image) but does not offer batching, page/size limits, or explicit advice for large-scale processing or resource management."},"maintainability":{"level":"Good","reason":"The content is modular with focused examples per task, a quick-reference table mapping tasks to tools, and parameterized examples (e.g., watermark options, pdf_operation XML), making it straightforward to adapt and reuse."}},"repo_name":"senweaver/senweaver-ide","skill_dir":"senweaver/senweaver-ide/.senweaver/skill/pdf"},{"skill_name":"pdf","skill_description":"Extracts text, tables, and structure from PDF files using opendataloader-pdf. Use when reading PDFs, converting PDF to markdown/JSON/HTML, extracting tables from PDF, processing scanned PDFs with OCR, or analyzing PDF content for RAG pipelines. Do NOT use for creating/editing PDFs, Word/Excel/PPT files (use xlsx2md for Excel), or viewing PDFs (use Read tool for simple PDF viewing).","relevance_score":55.1169,"author":"ssiumha","stars":9,"skill_url":"https://github.com/ssiumha/dots/blob/fb99c8b44ea48696a8e1b9983eb92cf6c3a14247/prompts/skills/pdf","category":"Research","evaluation":{"safety":{"level":"Good","reason":"The skill describes benign PDF extraction and explicitly notes sensitive-data handling (e.g., '--sanitize' and '민감 데이터 주의') and does not perform destructive actions; required powerful tools (Java/Python/Bash) are appropriate for the described CLI/API workflows."},"completeness":{"level":"Good","reason":"SKILL.md provides prerequisites (Java 11+, Python 3.10+), concrete CLI and Python API examples, a full options table (format, pages, hybrid, image/output, sanitize), hybrid-mode guidance, and usage principles, covering the essential steps and edge cases for converting PDFs."},"executability":{"level":"Good","reason":"The document gives exact commands (e.g., 'opendataloader-pdf <file.pdf> -o <output-dir> -f markdown'), a Python API snippet, and an options reference, so an agent with the listed tools can realistically run the workflow given the stated prerequisites."},"cost_awareness":{"level":"Good","reason":"The documentation is cost-conscious: it recommends batching to avoid repeated JVM startup, advises 'Fast mode first' and shows hybrid-mode routing so only complex pages go to the AI backend."},"maintainability":{"level":"Good","reason":"The skill is modular and configurable with many explicit options (e.g., '--format', '--hybrid', '--use-struct-tree') and clear principles like '배치 우선' that make it straightforward to adapt and reuse."}},"repo_name":"ssiumha/dots","skill_dir":"ssiumha/dots/prompts/skills/pdf"},{"skill_name":"pdf","skill_description":"Use this skill when generating PDF documents programmatically. Trigger when the user asks to \"create a PDF\", \"generate a PDF report\", \"export to PDF\", \"make a PDF with tables/charts\", or any task requiring structured PDF output with formatting, tables, or images.\n","relevance_score":54.8663,"author":"arcgolatkar","stars":0,"skill_url":"https://github.com/arcgolatkar/Skill_Collection/blob/8de026c708676b2f605238150774c2bbad36c86d/skills/document-creation/pdf","category":"Development","evaluation":{"safety":{"level":"Good","reason":"The skill is limited to generating PDF documents and includes only benign file operations (e.g., pdf.output, pdf.image with an os.path.exists check) and package installs; there are no destructive system commands or unsafe automation patterns in SKILL.md."},"completeness":{"level":"Good","reason":"The document includes prerequisites (pip install lines), step-by-step sections (Step 1–5) with code for titles, tables, headers/footers, images, common pitfalls, and a complete example, covering the essential actions needed to create PDFs with fpdf2."},"executability":{"level":"Good","reason":"Concrete, runnable Python snippets are provided (including pinned fpdf2==2.8.1) and demonstrate creating pages, tables, headers/footers, adding images, and saving the file, so an agent with a Python environment can execute the workflow as described."},"cost_awareness":{"level":"Good","reason":"PDF generation is a low-cost, local operation and the SKILL.md does not recommend expensive external services; prerequisites are lightweight Python packages (fpdf2/reportlab) appropriate for the task."},"maintainability":{"level":"Good","reason":"The skill is modular and reusable (e.g., custom ReportPDF class, clear sections, configurable col_widths and set_auto_page_break), making it straightforward to adapt or extend for other PDFs."}},"repo_name":"arcgolatkar/Skill_Collection","skill_dir":"arcgolatkar/Skill_Collection/skills/document-creation/pdf"},{"skill_name":"pdf","skill_description":"PDF 综合工具：提取文本与表格、生成/合并/拆分 PDF、处理表单。当需要填写 PDF 表单、批量处理/生成或分析 PDF 时使用。","relevance_score":54.7865,"author":"kunhai-88","stars":0,"skill_url":"https://github.com/kunhai-88/skills/blob/2742b5f777038de67de1f56d9c4694dd619313fa/pdf","category":"Development","evaluation":{"safety":{"level":"Good","reason":"PDF processing is a benign document operation domain with no destructive actions or security risks mentioned; the skill focuses on text/table extraction, merging, splitting, and form processing which are safe operations."},"completeness":{"level":"Average","reason":"The skill clearly identifies the goal (PDF processing) and mentions specific libraries (pypdf, pdfplumber, reportlab) with basic code examples, but assumes Python environment setup knowledge and references non-existent files (reference.md, forms.md) for advanced features."},"executability":{"level":"Good","reason":"The SKILL.md provides concrete, syntactically correct Python code snippets for core operations (reading, extracting text, merging, splitting) that an agent could follow; as an instruction-only skill for document processing, runnable scripts are not required."},"cost_awareness":{"level":"Good","reason":"PDF processing tasks described are inherently lightweight document operations with no heavy data processing, large-scale operations, or external API calls mentioned; the skill targets typical single-file or batch document workflows."},"maintainability":{"level":"Good","reason":"The skill has narrow, modular scope focused on PDF operations with clearly defined libraries (pypdf, pdfplumber, reportlab) and distinct capabilities (extract, merge, split, generate), making it easy to understand and adapt."}},"repo_name":"kunhai-88/skills","skill_dir":"kunhai-88/skills/pdf"}],"meta":{"query":"pdf","mode":"keyword","category":null,"limit":10,"total":4788,"page":1,"min_stars":0,"sort_by":"relevance","threshold":null}}