[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-langchain-ai-gpu-document-processing-tr":3,"guides-for-langchain-ai-gpu-document-processing":236,"similar-k177bv4rmjf0e7ypzrd861xcvn867ysw":237},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":22,"identity":191,"isFallback":196,"parentExtension":197,"providers":198,"relations":204,"repo":206,"workflow":232},1778053968286.4924,"k177bv4rmjf0e7ypzrd861xcvn867ysw",[],{"reviewCount":8},0,{"description":10,"name":11,"sourceUrl":12,"tags":13},"Use when processing large PDFs, document collections, or bulk text extraction tasks that benefit from GPU-accelerated processing. Triggers when the user provides large documents or needs bulk document analysis.","GPU Document Processing","https://github.com/langchain-ai/deepagents/tree/HEAD/examples/nvidia_deep_agent/skills/gpu-document-processing",[14,15,16,17,18,19,20,21],"gpu","document-processing","pdf","embedding","extraction","nvidia-nim","llm","agent",{"_creationTime":23,"_id":24,"extensionId":5,"locale":25,"result":26,"trustSignals":179,"workflow":189},1778054053159.3198,"kn73x8eqhx0vk6x8vkaqm107mh867yhf","en",{"checks":27,"evaluatedAt":169,"extensionSummary":170,"promptVersionExtension":171,"promptVersionScoring":172,"rationale":173,"score":174,"summary":175,"tags":176,"targetMarket":177,"tier":178},[28,33,36,39,43,46,50,55,58,61,65,69,72,76,79,82,85,88,91,94,98,102,106,110,114,117,120,123,128,131,134,137,140,143,147,150,153,156,159,162,166],{"category":29,"check":30,"severity":31,"summary":32},"Practical Utility","Problem relevance","pass","The description clearly states the problem addressed: processing large PDFs, document collections, and bulk text extraction, highlighting the benefit of GPU acceleration.",{"category":29,"check":34,"severity":31,"summary":35},"Unique selling proposition","The skill offers a unique selling proposition by offloading computationally intensive document processing tasks to a GPU sandbox, a distinct approach compared to standard CPU-based LLM reasoning.",{"category":29,"check":37,"severity":31,"summary":38},"Production readiness","The skill is production-ready, outlining a clear architecture (sandbox-as-tool), detailing capabilities like PDF text extraction, tabular data extraction, chunking, and embedding generation, and suggesting integration with NVIDIA NIM for scalable deployments.",{"category":40,"check":41,"severity":31,"summary":42},"Scope","Single responsibility principle","The skill focuses on GPU-accelerated document processing, including extraction, chunking, and embedding, which is a coherent and single domain.",{"category":40,"check":44,"severity":31,"summary":45},"Description quality","The displayed description accurately and concisely reflects the skill's capabilities and use cases.",{"category":47,"check":48,"severity":31,"summary":49},"Invocation","Scoped tools","The skill's capabilities are described as specific tasks like PDF text extraction, tabular data extraction, and document chunking, implying well-scoped operations rather than a single generalist tool.",{"category":51,"check":52,"severity":53,"summary":54},"Documentation","Configuration & parameter reference","not_applicable","No explicit configuration parameters or environment variables are documented, and the skill's implementation details suggest it operates based on the provided document references and internal defaults, making this check not applicable.",{"category":40,"check":56,"severity":53,"summary":57},"Tool naming","The skill's capabilities are described conceptually rather than as explicitly named tools. Therefore, evaluating tool naming directly is not applicable at this level.",{"category":40,"check":59,"severity":31,"summary":60},"Minimal I/O surface","The skill's capabilities are described in terms of processing specific document types and extracting structured data, implying focused inputs and outputs without extraneous information.",{"category":62,"check":63,"severity":31,"summary":64},"License","License usability","The repository contains a clear MIT license file, which is a permissive open-source license.",{"category":66,"check":67,"severity":53,"summary":68},"Maintenance","Commit recency","No commit history is available for evaluation.",{"category":66,"check":70,"severity":53,"summary":71},"Dependency Management","No third-party dependencies are explicitly managed or declared within the provided skill files, making this check not applicable.",{"category":73,"check":74,"severity":31,"summary":75},"Security","Secret Management","The skill's architecture explicitly states that API keys stay outside the sandbox, mitigating the risk of secrets being exposed during GPU processing.",{"category":73,"check":77,"severity":31,"summary":78},"Injection","The skill's architecture separates agent reasoning (CPU) from document processing (GPU sandbox), implying that data loaded into the sandbox for processing is treated as untrusted content.",{"category":73,"check":80,"severity":31,"summary":81},"Transitive Supply-Chain Grenades","The skill's architecture separates agent reasoning from GPU processing, and no remote fetching or execution of arbitrary code is apparent in the provided files.",{"category":73,"check":83,"severity":31,"summary":84},"Sandbox Isolation","The 'sandbox-as-tool' pattern implies that processing occurs in an isolated environment, with results returned to the agent, suggesting that operations are confined and do not affect the broader file system outside the skill's scope.",{"category":73,"check":86,"severity":31,"summary":87},"Sandbox escape primitives","No evidence of detached-process spawns or deny-retry loops is present in the provided skill description and architecture.",{"category":73,"check":89,"severity":31,"summary":90},"Data Exfiltration","The skill's architecture explicitly mentions that API keys stay outside the sandbox, and the focus is on processing documents, not exfiltrating confidential data.",{"category":73,"check":92,"severity":31,"summary":93},"Hidden Text Tricks","The provided skill files (SKILL.md) do not appear to contain any hidden text tricks or obfuscation methods.",{"category":95,"check":96,"severity":31,"summary":97},"Hooks","Opaque code execution","The skill description outlines an architecture but does not reveal any obfuscated code, base64 payloads, or runtime fetched scripts.",{"category":99,"check":100,"severity":31,"summary":101},"Portability","Structural Assumption","The skill describes a general document processing workflow and architecture, without making specific assumptions about user project organization or OS-specific paths.",{"category":103,"check":104,"severity":53,"summary":105},"Trust","Issues Attention","No issue data is available for evaluation.",{"category":107,"check":108,"severity":53,"summary":109},"Versioning","Release Management","No version information is present in the provided files, and no explicit install instructions that would necessitate versioning are detailed.",{"category":111,"check":112,"severity":53,"summary":113},"Code Execution","Validation","The provided skill description focuses on the architecture and capabilities, not specific input validation mechanisms for arguments or output sanitization. Thus, this check is not applicable without further code inspection.",{"category":73,"check":115,"severity":31,"summary":116},"Unguarded Destructive Operations","The skill focuses on document processing (extraction, analysis, embedding), which are not inherently destructive operations. No destructive primitives are mentioned.",{"category":111,"check":118,"severity":53,"summary":119},"Error Handling","Detailed error handling mechanisms are not described in the provided skill documentation; this would require code inspection.",{"category":111,"check":121,"severity":53,"summary":122},"Logging","Specific logging mechanisms for destructive actions or outbound calls are not detailed in the provided skill documentation; this would require code inspection.",{"category":124,"check":125,"severity":126,"summary":127},"Compliance","GDPR","info","The skill processes documents, which may contain personal data. While no specific sanitization is detailed, the architecture separates processing in a sandbox, which might limit direct exposure to the LLM without further context on data handling within the sandbox.",{"category":124,"check":129,"severity":31,"summary":130},"Target market","The skill's focus on document processing and integration with NVIDIA NIM suggests a global applicability for document analysis tasks, with no specific regional limitations apparent.",{"category":99,"check":132,"severity":31,"summary":133},"Runtime stability","The skill's architecture relies on a GPU sandbox and standard document processing tasks, which are likely to be stable across different runtime environments as long as the sandbox environment is correctly configured.",{"category":47,"check":135,"severity":31,"summary":136},"Precise Purpose","The description clearly defines the purpose (GPU-accelerated document processing) and triggers (large PDFs, bulk text extraction, embedding generation), and the architecture section clarifies the sandbox-as-tool pattern.",{"category":47,"check":138,"severity":31,"summary":139},"Concise Frontmatter","The frontmatter is concise and effectively summarizes the skill's purpose and use cases.",{"category":51,"check":141,"severity":31,"summary":142},"Concise Body","The SKILL.md body is well-structured and concise, outlining the architecture, capabilities, and workflow without excessive detail inline.",{"category":144,"check":145,"severity":53,"summary":146},"Context","Progressive Disclosure","The skill's documentation does not contain exceptionally long procedures or bulk third-party material that would necessitate progressive disclosure via separate files.",{"category":144,"check":148,"severity":53,"summary":149},"Forked exploration","The skill's primary function is document processing rather than deep exploration or code review, making the 'context: fork' setting not applicable.",{"category":29,"check":151,"severity":126,"summary":152},"Usage examples","While the skill describes its capabilities, specific end-to-end usage examples with inputs and expected outputs are not provided in the SKILL.md file.",{"category":29,"check":154,"severity":126,"summary":155},"Edge cases","The skill mentions handling multi-column layouts and corrupted pages for PDF extraction, but a comprehensive list of failure modes and recovery steps is not detailed.",{"category":111,"check":157,"severity":53,"summary":158},"Tool Fallback","The skill description does not mention reliance on external tools like MCP servers or custom environments, making this check not applicable.",{"category":99,"check":160,"severity":31,"summary":161},"Stack assumptions","The skill implies a need for a GPU environment and NVIDIA NIM for full functionality, which is a reasonable assumption for its described purpose.",{"category":163,"check":164,"severity":53,"summary":165},"Safety","Halt on unexpected state","The skill description does not detail specific preconditions or how unexpected states would be handled; this would require code inspection.",{"category":99,"check":167,"severity":31,"summary":168},"Cross-skill coupling","The skill is described as a standalone document processing capability and does not appear to implicitly rely on other skills being loaded.",1778054029810,"It employs a sandbox-as-tool architecture where complex processing is offloaded to a GPU environment while the agent handles reasoning on the CPU. The skill integrates with NVIDIA NIM for scalable deployments and is designed for tasks involving large documents or collections.","2.0.0","3.4.0","The extension is well-documented, clearly defines its purpose and architecture, and leverages a robust 'sandbox-as-tool' pattern for GPU-accelerated document processing. The MIT license is permissive, and security considerations like secret management and isolation are addressed. While usage examples and detailed edge case handling could be improved, the overall quality and utility are high.",85,"This skill provides GPU-accelerated document processing capabilities, including PDF text extraction, tabular data extraction, chunking, and embedding generation.",[14,15,16,17,18,19,20,21],"global","verified",{"codeQuality":180,"collectedAt":181,"documentation":182,"maintenance":184,"popularity":185,"security":186,"testCoverage":188},{},1778054018024,{"descriptionLength":183,"readmeSize":8},210,{},{"smitheryUniqueUsers":8,"smitheryUseCount":8},{"hasNpmPackage":187,"smitheryVerified":187},false,{"hasCi":187,"hasTests":187},{"updatedAt":190},1778054053159,{"githubOwner":192,"githubRepo":193,"locale":25,"slug":194,"type":195},"langchain-ai","deepagents","gpu-document-processing","skill",true,null,{"extract":199,"llm":202,"smithery":203},{"commitSha":200,"license":201},"b108c71d0c570e16c7050c1eac482e15dc35a5ed","MIT",{"promptVersionExtension":171,"promptVersionScoring":172,"score":174,"targetMarket":177,"tier":178},{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":187},{"repoId":205},"kd76dna2fvfbnjvzcpd2cwqnyd865xz7",{"_creationTime":207,"_id":205,"identity":208,"providers":210,"workflow":229},1777995558409.8704,{"githubOwner":192,"githubRepo":193,"sourceUrl":209},"https://github.com/langchain-ai/deepagents",{"discover":211,"github":215},{"sources":212},[213,214],"skills-sh","smithery",{"closedIssues90d":216,"forks":217,"homepage":218,"license":201,"openIssues90d":219,"pushedAt":220,"readmeSize":221,"stars":222,"topics":223},256,3140,"https://docs.langchain.com/deepagents",142,1778033560000,6232,22320,[193,224,225,226,227,228],"langchain","langgraph","ai","python","typescript",{"discoverAt":230,"extractAt":231,"githubAt":231,"updatedAt":231},1777995558409,1778053970345,{"anyEnrichmentAt":233,"extractAt":234,"githubAt":235,"llmAt":190,"smitheryAt":233,"updatedAt":190},1778053994907,1778053968286,1778053969344,[],[238,267,289,309,330,357],{"_creationTime":239,"_id":240,"community":241,"display":242,"identity":252,"providers":256,"relations":261,"workflow":263},1778053148350.4636,"k171dtxahnz3h8q0jz3gk6akks867ym1",{"reviewCount":8},{"description":243,"installMethods":244,"name":245,"sourceUrl":246,"tags":247},"Extract text, tables, and metadata from PDFs using pdfplumber",{},"PDF Extraction","https://github.com/claude-office-skills/skills/tree/HEAD/pdf-extraction",[16,18,248,249,250,251,15],"text","tables","metadata","pdfplumber",{"githubOwner":253,"githubRepo":254,"locale":25,"slug":255,"type":195},"claude-office-skills","skills","pdf-extraction",{"extract":257,"llm":259},{"commitSha":258,"license":201},"9c4c7d5cd2813a8936bf2c9fdb174ea883b85a11",{"promptVersionExtension":171,"promptVersionScoring":172,"score":260,"targetMarket":177,"tier":178},95,{"repoId":262},"kd7fw7xbj58qc2z8whrrjptbed8659db",{"anyEnrichmentAt":264,"extractAt":265,"githubAt":264,"llmAt":266,"updatedAt":266},1778053151766,1778053148350,1778053561145,{"_creationTime":268,"_id":269,"community":270,"display":271,"identity":281,"providers":283,"relations":287,"workflow":288},1778053148350.4373,"k1776t2fdx4h35mkwpc5h201dd866zms",{"reviewCount":8},{"description":272,"installMethods":273,"name":274,"sourceUrl":275,"tags":276},">",{},"Document Parser Skill","https://github.com/claude-office-skills/skills/tree/HEAD/doc-parser",[277,15,16,227,278,18,279,280],"parsing","ocr","layout-analysis","docling",{"githubOwner":253,"githubRepo":254,"locale":25,"slug":282,"type":195},"doc-parser",{"extract":284,"llm":285},{"commitSha":258,"license":201},{"promptVersionExtension":171,"promptVersionScoring":172,"score":286,"targetMarket":177,"tier":178},92,{"repoId":262},{"anyEnrichmentAt":264,"extractAt":265,"githubAt":264,"llmAt":266,"updatedAt":266},{"_creationTime":290,"_id":291,"community":292,"display":293,"identity":301,"providers":303,"relations":307,"workflow":308},1778053148350.4656,"k171nxqak0bb4qq89mkfwf02s5867cf6",{"reviewCount":8},{"description":294,"installMethods":295,"name":296,"sourceUrl":297,"tags":298},"Convert PDF files to editable Word documents using pdf2docx",{},"PDF to DOCX Converter","https://github.com/claude-office-skills/skills/tree/HEAD/pdf-to-docx",[16,299,300,15,227],"docx","conversion",{"githubOwner":253,"githubRepo":254,"locale":25,"slug":302,"type":195},"pdf-to-docx",{"extract":304,"llm":305},{"commitSha":258,"license":201},{"promptVersionExtension":171,"promptVersionScoring":172,"score":306,"targetMarket":177,"tier":178},98,{"repoId":262},{"anyEnrichmentAt":264,"extractAt":265,"githubAt":264,"llmAt":266,"updatedAt":266},{"_creationTime":310,"_id":311,"community":312,"display":313,"identity":323,"providers":325,"relations":328,"workflow":329},1778053148350.4265,"k171agyyd8nv26rt447dvhy0998669wm",{"reviewCount":8},{"description":314,"installMethods":315,"name":316,"sourceUrl":317,"tags":318},"Answer questions about PDF content, summarize, and extract information",{},"Chat with PDF","https://github.com/claude-office-skills/skills/tree/HEAD/chat-with-pdf",[16,319,18,320,321,322],"qa","summarization","mcp","documentation",{"githubOwner":253,"githubRepo":254,"locale":25,"slug":324,"type":195},"chat-with-pdf",{"extract":326,"llm":327},{"commitSha":258,"license":201},{"promptVersionExtension":171,"promptVersionScoring":172,"score":260,"targetMarket":177,"tier":178},{"repoId":262},{"anyEnrichmentAt":264,"extractAt":265,"githubAt":264,"llmAt":266,"updatedAt":266},{"_creationTime":331,"_id":332,"community":333,"display":334,"identity":343,"providers":347,"relations":351,"workflow":353},1778054086261.0896,"k173qc4w9k717e5jcb4da4zrcx866a1a",{"reviewCount":8},{"description":335,"name":336,"sourceUrl":337,"tags":338},"Recognize, diagnose, and mitigate patterns of context degradation in agent systems. Use when context grows large, agent performance degrades unexpectedly, or debugging agent failures.","Context Degradation Patterns","https://github.com/bilalmk/todo_correct/tree/HEAD/.claude/skills/mjs/context-degradation",[21,339,340,341,20,342],"context","performance","debugging","analysis",{"githubOwner":344,"githubRepo":345,"locale":25,"slug":346,"type":195},"bilalmk","todo_correct","context-degradation",{"extract":348,"llm":350},{"commitSha":349,"license":201},"8b43aa04bd5c53e3cda46469b953684519a84ea7",{"promptVersionExtension":171,"promptVersionScoring":172,"score":260,"targetMarket":177,"tier":178},{"repoId":352},"kd75ecf652eb91ha327s8bqbex865z6v",{"anyEnrichmentAt":354,"extractAt":355,"githubAt":354,"llmAt":356,"updatedAt":356},1778054086910,1778054086261,1778054163453,{"_creationTime":358,"_id":359,"community":360,"display":361,"identity":369,"providers":371,"relations":374,"workflow":375},1778053148350.4734,"k1782aqmjfqy0qysysgq76w9z1867e3x",{"reviewCount":8},{"description":272,"installMethods":362,"name":363,"sourceUrl":364,"tags":365},{},"Smart OCR Skill","https://github.com/claude-office-skills/skills/tree/HEAD/smart-ocr",[278,277,366,367,18,368,15],"multilingual","paddleocr","image-processing",{"githubOwner":253,"githubRepo":254,"locale":25,"slug":370,"type":195},"smart-ocr",{"extract":372,"llm":373},{"commitSha":258,"license":201},{"promptVersionExtension":171,"promptVersionScoring":172,"score":286,"targetMarket":177,"tier":178},{"repoId":262},{"anyEnrichmentAt":264,"extractAt":265,"githubAt":264,"llmAt":266,"updatedAt":266}]