@byted-las/contextlake-openclaw 1.0.0 → 1.0.3

package/dist/src/commands/tools.d.ts

@@ -1,219 +1,11 @@
-export declare function getAgentTools(pluginConfig: any, logger: any): {
-    ingestTool: {
-        name: string;
-        label: string;
-        description: string;
-        parameters: {
-            type: string;
-            properties: {
-                files: {
-                    type: string;
-                    items: {
-                        type: string;
-                    };
-                    description: string;
-                };
-                metadata: {
-                    type: string;
-                    description: string;
-                    additionalProperties: boolean;
-                };
-                chunkSize: {
-                    type: string;
-                    description: string;
-                };
-                overlap: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: string[];
-            additionalProperties: boolean;
-        };
-        schema: {
-            type: string;
-            properties: {
-                files: {
-                    type: string;
-                    items: {
-                        type: string;
-                    };
-                    description: string;
-                };
-                metadata: {
-                    type: string;
-                    description: string;
-                    additionalProperties: boolean;
-                };
-                chunkSize: {
-                    type: string;
-                    description: string;
-                };
-                overlap: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: string[];
-            additionalProperties: boolean;
-        };
-        execute(toolCallId: string, params: any): Promise<{
-            success: boolean;
-            result: any;
-            error?: undefined;
-        } | {
-            success: boolean;
-            error: any;
-            result?: undefined;
-        }>;
-    };
-    retrieveTool: {
-        name: string;
-        label: string;
-        description: string;
-        parameters: {
-            type: string;
-            properties: {
-                query: {
-                    type: string;
-                    description: string;
-                };
-                top_k: {
-                    type: string;
-                    description: string;
-                };
-                filter: {
-                    type: string;
-                    description: string;
-                };
-                include_binary: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: string[];
-            additionalProperties: boolean;
-        };
-        schema: {
-            type: string;
-            properties: {
-                query: {
-                    type: string;
-                    description: string;
-                };
-                top_k: {
-                    type: string;
-                    description: string;
-                };
-                filter: {
-                    type: string;
-                    description: string;
-                };
-                include_binary: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: string[];
-            additionalProperties: boolean;
-        };
-        execute(toolCallId: string, params: any): Promise<{
-            success: boolean;
-            result: any[];
-            error?: undefined;
-        } | {
-            success: boolean;
-            error: any;
-            result?: undefined;
-        }>;
-    };
-    listTool: {
-        name: string;
-        label: string;
-        description: string;
-        parameters: {
-            type: string;
-            properties: {
-                limit: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: never[];
-            additionalProperties: boolean;
-        };
-        schema: {
-            type: string;
-            properties: {
-                limit: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: never[];
-        };
-        execute(toolCallId: string, params: any): Promise<{
-            success: boolean;
-            result: any;
-            error?: undefined;
-        } | {
-            success: boolean;
-            error: any;
-            result?: undefined;
-        }>;
-    };
-    deleteTool: {
-        name: string;
-        label: string;
-        description: string;
-        parameters: {
-            type: string;
-            properties: {
-                file_ids: {
-                    type: string;
-                    items: {
-                        type: string;
-                    };
-                    description: string;
-                };
-                filter: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: never[];
-            additionalProperties: boolean;
-        };
-        schema: {
-            type: string;
-            properties: {
-                file_ids: {
-                    type: string;
-                    items: {
-                        type: string;
-                    };
-                    description: string;
-                };
-                filter: {
-                    type: string;
-                    description: string;
-                };
-            };
-            required: never[];
-        };
-        execute(toolCallId: string, params: any): Promise<{
-            success: boolean;
-            result: {
-                status: string;
-                message: string;
-                deleted_count: number;
-                storage_cleaned: number;
-            };
-            error?: undefined;
-        } | {
-            success: boolean;
-            error: any;
-            result?: undefined;
-        }>;
-    };
+import { ContextLakeConfig } from '../utils/config';
+import type { AnyAgentTool } from 'openclaw/plugin-sdk';
+export declare function getAgentTools(pluginConfig: ContextLakeConfig, logger: any): {
+    ingestTool: AnyAgentTool;
+    retrieveTool: AnyAgentTool;
+    listTool: AnyAgentTool;
+    deleteTool: AnyAgentTool;
+    lasDataProfilerTool: AnyAgentTool;
+    listDatasourceTool: AnyAgentTool;
+    lasTools: AnyAgentTool[];
 };

package/dist/src/commands/tools.js

@@ -4,69 +4,53 @@ exports.getAgentTools = getAgentTools;
 const ingest_1 = require("../lib/actions/ingest");
 const retrieve_1 = require("../lib/actions/retrieve");
 const manage_1 = require("../lib/actions/manage");
+const profiler_1 = require("../lib/actions/profiler");
+const las_tools_1 = require("../lib/actions/las-tools");
 function getAgentTools(pluginConfig, logger) {
+    const lasTools = (0, las_tools_1.getLasTools)(pluginConfig, logger);
     return {
+        lasTools,
+        listDatasourceTool: {
+            name: 'contextlake-list-datasource',
+            label: 'ContextLake List Datasources',
+            description: `List all connected and profiled data sources.`,
+            parameters: {
+                type: 'object',
+                properties: {},
+                required: [],
+                additionalProperties: false
+            },
+            async execute(toolCallId, params) {
+                logger.info(`[${new Date().toISOString()}] [ContextLake] Executing list-datasource skill, toolCallId: ${toolCallId}`);
+                try {
+                    const result = await (0, profiler_1.listDataSources)();
+                    return {
+                        content: [{ type: "text", text: JSON.stringify(result) }],
+                        details: result
+                    };
+                }
+                catch (error) {
+                    logger.error(`[${new Date().toISOString()}] [ContextLake] list-datasource skill failed`, { error: error.message });
+                    return {
+                        content: [{ type: "text", text: String(error.message) }],
+                        details: { error: error.message }
+                    };
+                }
+            }
+        },
         ingestTool: {
             name: 'contextlake-ingest',
             label: 'ContextLake Ingest',
-            description: `Upload, ingest, and index documents into the ContextLake Knowledge Base (知识库) / Knowledge Lake (知识湖).
+            description: `Process and ingest all files from a connected data source into the knowledge base.
 Use this tool when the user wants to "将知识注入", "上传文件", "入库", "添加文档", "ingest files", or "add knowledge".
-Supports processing of various file types including PDF, Word, Markdown, and Text.
-Automatically handles text extraction, cleaning, chunking, embedding generation, and storage.
-
-Example User Queries:
-- "帮我把这个文档注入到知识湖中"
-- "上传这份 PDF 到知识库"
-- "Please ingest these documents into ContextLake"
-- "将 /path/to/doc.txt 添加到知识库"`,
+Supports multimodal files (text, images, audio, video, pdf) by using LAS models to understand and embed them.
+Must be called after a data source has been successfully profiled via \`las-data-profiler\`.`,
             parameters: {
                 type: 'object',
                 properties: {
-                    files: {
-                        type: 'array',
-                        items: { type: 'string' },
-                        description: 'List of file paths to ingest'
-                    },
-                    metadata: {
-                        type: 'object',
-                        description: 'Optional JSON metadata to attach to documents',
-                        additionalProperties: true
-                    },
-                    chunkSize: {
-                        type: 'integer',
-                        description: 'Chunk size for text splitting'
-                    },
-                    overlap: {
-                        type: 'integer',
-                        description: 'Overlap size for text splitting'
-                    }
+                    datasource_name: { type: 'string', description: 'Name of the data source previously profiled' }
                 },
-                required: ['files'],
-                additionalProperties: false
-            },
-            schema: {
-                type: 'object',
-                properties: {
-                    files: {
-                        type: 'array',
-                        items: { type: 'string' },
-                        description: 'List of file paths to ingest'
-                    },
-                    metadata: {
-                        type: 'object',
-                        description: 'Optional JSON metadata to attach to documents',
-                        additionalProperties: true
-                    },
-                    chunkSize: {
-                        type: 'integer',
-                        description: 'Chunk size for text splitting'
-                    },
-                    overlap: {
-                        type: 'integer',
-                        description: 'Overlap size for text splitting'
-                    }
-                },
-                required: ['files'],
+                required: ['datasource_name'],
                 additionalProperties: false
             },
             async execute(toolCallId, params) {
@@ -80,32 +64,32 @@ Example User Queries:
                         catch (e) {
                             logger.warn(`[ContextLake] Received string params, possibly toolCallId?`, { params });
                             return {
-                                success: false,
-                                error: `Invalid params format: received string "${params}", expected object with 'files' array.`
+                                content: [{ type: "text", text: `Invalid params format: received string "${params}", expected object with 'datasource_name'.` }],
+                                details: { error: `Invalid params format: received string "${params}", expected object with 'datasource_name'.` }
                             };
                         }
                     }
-                    if (!actualParams.files && actualParams.params && actualParams.params.files) {
+                    if (!actualParams.datasource_name && actualParams.params && actualParams.params.datasource_name) {
                         actualParams = actualParams.params;
                     }
-                    if (!actualParams.files || !Array.isArray(actualParams.files)) {
+                    if (!actualParams.datasource_name) {
                         return {
-                            success: false,
-                            error: `Invalid params: 'files' must be an array. Received keys: ${Object.keys(actualParams)}`
+                            content: [{ type: "text", text: `Invalid params: 'datasource_name' is required. Received keys: ${Object.keys(actualParams)}` }],
+                            details: { error: `Invalid params: 'datasource_name' is required. Received keys: ${Object.keys(actualParams)}` }
                         };
                     }
-                    const result = await (0, ingest_1.ingestAssets)(actualParams, pluginConfig, logger);
+                    const result = await (0, ingest_1.ingestSource)(actualParams, pluginConfig, logger);
                     logger.info(`[${new Date().toISOString()}] [ContextLake] Ingest skill completed successfully`, { resultSummary: Array.isArray(result) ? `Processed ${result.length} items` : 'Success' });
                     return {
-                        success: true,
-                        result
+                        content: [{ type: "text", text: JSON.stringify(result) }],
+                        details: result
                     };
                 }
                 catch (error) {
                     logger.error(`[${new Date().toISOString()}] [ContextLake] Ingest skill failed`, { error: error.message, stack: error.stack });
                     return {
-                        success: false,
-                        error: error.message
+                        content: [{ type: "text", text: String(error.message) }],
+                        details: { error: error.message }
                     };
                 }
             }
@@ -133,17 +117,6 @@ Example User Queries:
                 required: ['query'],
                 additionalProperties: false
             },
-            schema: {
-                type: 'object',
-                properties: {
-                    query: { type: 'string', description: 'Search query' },
-                    top_k: { type: 'integer', description: 'Number of results to return' },
-                    filter: { type: 'string', description: 'Filter string' },
-                    include_binary: { type: 'boolean', description: 'Whether to include binary content' }
-                },
-                required: ['query'],
-                additionalProperties: false
-            },
             async execute(toolCallId, params) {
                 logger.info(`[${new Date().toISOString()}] [ContextLake] Executing retrieve skill, toolCallId: ${toolCallId}`, { params: JSON.stringify(params) });
                 try {
@@ -163,22 +136,23 @@ Example User Queries:
                     }
                     if (!actualParams || typeof actualParams.query !== 'string') {
                         return {
-                            success: false,
-                            error: `Invalid params: 'query' is required and must be a string. Received: ${JSON.stringify(actualParams)}`
+                            content: [{ type: "text", text: `Invalid params: 'query' is required and must be a string. Received: ${JSON.stringify(actualParams)}` }],
+                            details: { error: `Invalid params: 'query' is required and must be a string. Received: ${JSON.stringify(actualParams)}` }
                         };
                     }
                     const result = await (0, retrieve_1.retrieveAssets)(actualParams, pluginConfig, logger);
                     logger.info(`[${new Date().toISOString()}] [ContextLake] Retrieve skill completed`, { resultCount: Array.isArray(result) ? result.length : 0 });
                     return {
-                        success: true,
-                        result
+                        content: [{ type: "text", text: JSON.stringify(result) }],
+                        details: result
                     };
                 }
                 catch (error) {
                     logger.error(`[${new Date().toISOString()}] [ContextLake] Retrieve skill failed`, { error: error.message, stack: error.stack });
                     return {
-                        success: false,
-                        error: error.message
+                        content: [{ type: "text", text: String(error.message) }],
+                        details: { error: error.message
+                        }
                     };
                 }
             }
@@ -201,13 +175,6 @@ Example User Queries:
                 required: [],
                 additionalProperties: false
             },
-            schema: {
-                type: 'object',
-                properties: {
-                    limit: { type: 'integer', description: 'Limit for list action' }
-                },
-                required: []
-            },
             async execute(toolCallId, params) {
                 logger.info(`[${new Date().toISOString()}] [ContextLake] Executing list skill, toolCallId: ${toolCallId}`, { params: JSON.stringify(params) });
                 try {
@@ -218,15 +185,16 @@ Example User Queries:
                     const result = await (0, manage_1.listAssets)(actualParams, pluginConfig, logger);
                     logger.info(`[${new Date().toISOString()}] [ContextLake] List skill completed`, { count: Array.isArray(result) ? result.length : 0 });
                     return {
-                        success: true,
-                        result
+                        content: [{ type: "text", text: JSON.stringify(result) }],
+                        details: result
                     };
                 }
                 catch (error) {
                     logger.error(`[${new Date().toISOString()}] [ContextLake] List skill failed`, { error: error.message, stack: error.stack });
                     return {
-                        success: false,
-                        error: error.message
+                        content: [{ type: "text", text: String(error.message) }],
+                        details: { error: error.message
+                        }
                     };
                 }
             }
@@ -251,14 +219,6 @@ Example User Queries:
                 required: [],
                 additionalProperties: false
             },
-            schema: {
-                type: 'object',
-                properties: {
-                    file_ids: { type: 'array', items: { type: 'string' }, description: 'File IDs to delete' },
-                    filter: { type: 'string', description: 'Filter string for deletion' }
-                },
-                required: []
-            },
             async execute(toolCallId, params) {
                 logger.info(`[${new Date().toISOString()}] [ContextLake] Executing delete skill, toolCallId: ${toolCallId}`, { params: JSON.stringify(params) });
                 try {
@@ -269,15 +229,60 @@ Example User Queries:
                     const result = await (0, manage_1.deleteAssets)(actualParams, pluginConfig, logger);
                     logger.info(`[${new Date().toISOString()}] [ContextLake] Delete skill completed`, { result });
                     return {
-                        success: true,
-                        result
+                        content: [{ type: "text", text: JSON.stringify(result) }],
+                        details: result
                     };
                 }
                 catch (error) {
                     logger.error(`[${new Date().toISOString()}] [ContextLake] Delete skill failed`, { error: error.message, stack: error.stack });
                     return {
-                        success: false,
-                        error: error.message
+                        content: [{ type: "text", text: String(error.message) }],
+                        details: { error: error.message
+                        }
+                    };
+                }
+            }
+        },
+        lasDataProfilerTool: {
+            name: 'las-data-profiler',
+            label: 'LAS Data Profiler',
+            description: 'Connect to a data source (TOS/OSS/COS/S3/Local) and profile its structure, schemas, and media metadata into LanceDB',
+            parameters: {
+                type: 'object',
+                properties: {
+                    datasource_name: { type: 'string', description: 'Name of the data source' },
+                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'], description: 'Data source type' },
+                    endpoint: { type: 'string', description: 'S3 Endpoint URL (not needed for local)' },
+                    access_key: { type: 'string', description: 'Credential ID for the data source' },
+                    secret_key: { type: 'string', description: 'Credential value for the data source' },
+                    region: { type: 'string', description: 'Region identifier (e.g. cn-beijing)' },
+                    bucket: { type: 'string', description: 'Bucket name (or local root directory for local vendor)' },
+                    prefix: { type: 'string', description: 'Path prefix to limit scan scope' },
+                    sample_rows: { type: 'integer', description: 'Number of rows to sample per structured file' }
+                },
+                required: ['datasource_name', 'vendor', 'bucket', 'prefix'],
+                additionalProperties: false
+            },
+            async execute(toolCallId, params) {
+                logger.info(`[${new Date().toISOString()}] [ContextLake] Executing las-data-profiler skill, toolCallId: ${toolCallId}`, { params: JSON.stringify(params) });
+                try {
+                    let actualParams = params;
+                    if (params && params.params) {
+                        actualParams = params.params;
+                    }
+                    const result = await (0, profiler_1.connectDataSource)(actualParams);
+                    logger.info(`[${new Date().toISOString()}] [ContextLake] las-data-profiler skill completed`, { result });
+                    return {
+                        content: [{ type: "text", text: JSON.stringify(result) }],
+                        details: result
+                    };
+                }
+                catch (error) {
+                    logger.error(`[${new Date().toISOString()}] [ContextLake] las-data-profiler skill failed`, { error: error.message, stack: error.stack });
+                    return {
+                        content: [{ type: "text", text: String(error.message) }],
+                        details: { error: error.message
+                        }
                     };
                 }
             }

package/dist/src/lib/actions/ingest-source.d.ts

@@ -0,0 +1,15 @@
+import { ContextLakeConfig } from '../../utils/config';
+export interface IngestSourceParams {
+    datasource_name: string;
+}
+export declare function ingestSource(params: IngestSourceParams, config: ContextLakeConfig, logger?: any): Promise<({
+    file: any;
+    status: string;
+    chunks: number;
+    message?: undefined;
+} | {
+    file: any;
+    status: string;
+    message: any;
+    chunks?: undefined;
+})[]>;