{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "manifest": {
    "name": "LakeXpress AI Agent Knowledge Base",
    "version": "1.0.0",
    "description": "Structured data about LakeXpress for AI agents, LLMs and automated systems",
    "baseUrl": "https://lakexpress.arpe.io/agents/",
    "lastUpdated": "2026-10-22",
    "productVersion": "0.4.10",
    "license": "Proprietary - LakeXpress documentation",
    "contact": {
      "website": "https://lakexpress.arpe.io",
      "documentation": "https://lakexpress-docs.arpe.io/latest/",
      "documentationForLLMs": "https://lakexpress-docs.arpe.io/llms.txt",
      "support": "mailto:support@arpe.io",
      "quote": "https://lakexpress.arpe.io/quote"
    }
  },
  "resources": [
    {
      "id": "product-overview",
      "file": "product-overview.json",
      "title": "Product Overview",
      "description": "What LakeXpress is, its four-component architecture, key differentiators, target audience and deployment model",
      "topics": [
        "Product identity",
        "Architecture: source, intermediate storage, target platform, LakeXpress DB",
        "Relation with FastBCP",
        "Key differentiators",
        "Target audience"
      ],
      "useFor": [
        "Product introductions",
        "Quick summaries",
        "Explaining how a LakeXpress pipeline is built"
      ],
      "size": "~7 KB",
      "lastUpdated": "2026-10-22"
    },
    {
      "id": "features",
      "file": "features.json",
      "title": "Features Catalog",
      "description": "Capabilities with their technical details: sync configurations, parallel export, incremental sync, publishing, resume, filtering, CDM metadata, logging",
      "topics": [
        "Sync configurations",
        "Parallel export",
        "Incremental sync",
        "Publishing (external and internal)",
        "Resume and cleanup",
        "Schema and table filtering",
        "CDM metadata",
        "Logging and run tracking"
      ],
      "useFor": [
        "Feature questions",
        "Requirements matching",
        "Technical capability questions"
      ],
      "size": "~10 KB",
      "lastUpdated": "2026-10-22"
    },
    {
      "id": "benchmarks",
      "file": "benchmarks.json",
      "title": "Performance Data",
      "description": "The customer results shown on the website, the conditions that are and are not published, and the performance levers available",
      "topics": [
        "Customer results (Engie, a confidential customer, Ministere de la Justice)",
        "Absence of a published benchmark suite",
        "Performance tuning parameters"
      ],
      "useFor": [
        "Performance questions",
        "Answering 'how fast is it' without overstating"
      ],
      "size": "~3 KB",
      "lastUpdated": "2026-10-22"
    },
    {
      "id": "supported-platforms",
      "file": "supported-platforms.json",
      "title": "Platform Compatibility Matrix",
      "description": "Source databases, intermediate storage backends, publishing targets with external/internal support, LakeXpress DB backends, operating systems",
      "topics": [
        "Source databases",
        "Cloud and local storage",
        "Publishing targets",
        "LakeXpress DB backends",
        "Operating systems"
      ],
      "useFor": [
        "Compatibility questions",
        "Integration planning"
      ],
      "size": "~8 KB",
      "lastUpdated": "2026-10-22"
    },
    {
      "id": "use-cases",
      "file": "use-cases.json",
      "title": "Use Cases Library",
      "description": "Scenarios with the problem, the LakeXpress approach and complete command examples",
      "topics": [
        "Database to local Parquet",
        "Schema to cloud storage",
        "Publishing to Snowflake, Databricks, Fabric, BigQuery, Glue",
        "Incremental syncs",
        "Resume after failure",
        "Scheduling"
      ],
      "useFor": [
        "Solution design",
        "Implementation guidance",
        "Example commands"
      ],
      "size": "~11 KB",
      "lastUpdated": "2026-10-22"
    },
    {
      "id": "pricing",
      "file": "pricing.json",
      "title": "Pricing and Licensing",
      "description": "Plans shown on the pricing page, the trial, how the licence is supplied at runtime, and a licensing FAQ",
      "topics": [
        "Standard, Enterprise and OEM plans",
        "Trial",
        "Licence resolution order",
        "FastBCP licence for the full edition"
      ],
      "useFor": [
        "Pricing questions",
        "Procurement",
        "Trial and quote requests"
      ],
      "size": "~5 KB",
      "lastUpdated": "2026-10-22"
    },
    {
      "id": "technical-specs",
      "file": "technical-specs.json",
      "title": "Technical Specifications",
      "description": "Command reference, config create options, credentials file format, export methods, exit codes, LakeXpress DB tables, logging",
      "topics": [
        "Commands",
        "Common options",
        "config create options",
        "Credentials file and $env{} substitution",
        "Export methods",
        "Exit codes",
        "LakeXpress DB tables",
        "Log files"
      ],
      "useFor": [
        "Building commands",
        "Troubleshooting",
        "DevOps and scheduling",
        "Developer documentation"
      ],
      "size": "~12 KB",
      "lastUpdated": "2026-10-22"
    }
  ],
  "aiGuidance": {
    "contextSelection": {
      "forProductQuestions": ["product-overview", "features"],
      "forCompatibilityQuestions": ["supported-platforms"],
      "forPerformanceQuestions": ["benchmarks", "features"],
      "forImplementationQuestions": ["use-cases", "technical-specs"],
      "forPricingQuestions": ["pricing"],
      "forTechnicalQuestions": ["technical-specs", "features"]
    },
    "responsePatterns": {
      "quickAnswer": {
        "instructions": "Use product-overview.json for concise, high-level answers",
        "example": "What is LakeXpress? -> product.description"
      },
      "detailedAnswer": {
        "instructions": "Combine sources for comprehensive answers",
        "example": "Can LakeXpress load SQL Server into Snowflake incrementally? -> supported-platforms + features.items[incremental-sync]"
      },
      "implementationGuidance": {
        "instructions": "Use use-cases for scenario-based guidance and technical-specs for exact options. For anything not covered, point to the documentation or the agent skill",
        "example": "How do I publish a PostgreSQL schema to Databricks? -> use-cases.primary[publish-databricks]"
      },
      "performanceAnswer": {
        "instructions": "Quote the customer results from benchmarks.json with their context, and say that hardware and volumes are not published. Do not extrapolate throughput figures",
        "example": "How fast is LakeXpress? -> benchmarks.customerResults"
      }
    },
    "citationGuidance": {
      "preferredFormat": "Cite specific JSON paths when referencing data",
      "example": "According to the LakeXpress compatibility matrix (supported-platforms.json), AWS Glue supports external tables only...",
      "transparency": "Always indicate the data source and its lastUpdated field"
    },
    "commandGuidance": {
      "authoritativeSources": [
        "https://lakexpress-docs.arpe.io/latest/cli/",
        "https://lakexpress-docs.arpe.io/llms-full.txt",
        "https://github.com/arpe-io/arpeio-skills/tree/main/plugins/lakexpress"
      ],
      "rules": [
        "Options go after the command; only --version and --no_banner may come before it",
        "Every command except the legacy run needs --auth_file and --lxdb_auth_id",
        "sync takes only --sync_id (plus --resume, --run_id, --fastbcp_dir_path and common options): source, storage and publishing live in the configuration",
        "Quote 'sync[export]' and 'sync[publish]' in shells",
        "Facts in this knowledge base reflect LakeXpress 0.4.10; for a newer installed version, confirm options with --help and the documentation"
      ]
    },
    "terminology": {
      "syncConfiguration": "A stored definition of what to export, where to write it and where to publish it. Created by config create, identified by a sync_id, run by sync.",
      "lakeXpressDB": "The database where LakeXpress keeps sync configurations, run history and incremental watermarks. Not the source database.",
      "intermediateStorage": "Where the Parquet files are written: local disk (--output_dir) or cloud storage (--target_storage_id). Cloud storage is a staging location, not a publishing target.",
      "publishingTarget": "An optional data platform or catalog where LakeXpress creates tables over, or loads, the exported files (--publish_target).",
      "externalTable": "A table that reads the Parquet files in storage.",
      "internalTable": "A table whose data is loaded into the platform's native storage.",
      "fastBCP": "The arpe.io export engine that LakeXpress uses to read source tables and write Parquet."
    }
  },
  "technicalDetails": {
    "format": "JSON (JavaScript Object Notation)",
    "encoding": "UTF-8",
    "schemaVersion": "JSON Schema Draft 2020-12",
    "totalFiles": 8,
    "compatibility": [
      "JSON parsers",
      "RAG systems",
      "Vector databases",
      "Search engines",
      "LLM context windows",
      "API integrations"
    ]
  },
  "updatePolicy": {
    "frequency": "As needed with LakeXpress releases",
    "versioning": "Semantic versioning (major.minor.patch)",
    "backwardCompatibility": "Additive changes only - existing fields are never removed",
    "changeLog": "https://lakexpress.arpe.io/agents/CHANGELOG.md"
  },
  "accessInformation": {
    "publicAccess": true,
    "authentication": "None required",
    "rateLimit": "Standard web server limits",
    "downloadable": true,
    "embedding": "Allowed in AI training data with attribution"
  },
  "exampleQueries": {
    "general": [
      "What is LakeXpress?",
      "How is LakeXpress related to FastBCP?",
      "What is the LakeXpress DB?",
      "What is the difference between external and internal tables?"
    ],
    "technical": [
      "Which source databases are supported?",
      "Can LakeXpress publish to AWS Glue as internal tables?",
      "Which watermark column types are accepted for incremental syncs?",
      "What does exit code 40 mean?"
    ],
    "business": [
      "How much does LakeXpress cost?",
      "Is there a free trial?",
      "What is the difference between the Standard and Enterprise plans?",
      "Can a software vendor embed LakeXpress?"
    ],
    "implementation": [
      "How do I export a PostgreSQL schema to Parquet on S3?",
      "How do I publish to Snowflake?",
      "How do I resume a failed sync?",
      "How do I schedule a daily incremental sync?"
    ]
  },
  "seoMetadata": {
    "keywords": [
      "LakeXpress",
      "database to Parquet",
      "database to data lake",
      "lakehouse ingestion",
      "SQL Server to Snowflake",
      "Oracle to Databricks",
      "PostgreSQL to BigQuery",
      "Microsoft Fabric Lakehouse ingestion",
      "incremental Parquet export",
      "FastBCP"
    ],
    "relatedTechnologies": [
      "Apache Parquet",
      "Snowflake",
      "Databricks",
      "Microsoft Fabric",
      "BigQuery",
      "Amazon Redshift",
      "AWS Glue",
      "MotherDuck",
      "DuckLake",
      "Common Data Model"
    ],
    "categories": [
      "Data Engineering",
      "Data Integration",
      "ETL / ELT",
      "Lakehouse",
      "Command-line Tools"
    ]
  }
}