@mcp-graph-workflow/mcp-graph 3.0.0

package/dist/core/rag/capture-indexer.d.ts

@@ -0,0 +1,16 @@
+/**
+ * Capture Indexer — indexes web capture results into the knowledge store.
+ * Called after captureWebPage() to persist extracted content for search.
+ */
+import type { CaptureResult } from "../capture/web-capture.js";
+import { KnowledgeStore } from "../store/knowledge-store.js";
+export interface CaptureIndexResult {
+    documentsIndexed: number;
+    url: string;
+}
+/**
+ * Index a web capture result into the knowledge store.
+ * Chunks the extracted text and stores with source_type='web_capture'.
+ */
+export declare function indexCapture(knowledgeStore: KnowledgeStore, capture: CaptureResult): CaptureIndexResult;
+//# sourceMappingURL=capture-indexer.d.ts.map

package/dist/core/rag/capture-indexer.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"capture-indexer.d.ts","sourceRoot":"","sources":["../../../src/core/rag/capture-indexer.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,2BAA2B,CAAC;AAC/D,OAAO,EAAE,cAAc,EAAE,MAAM,6BAA6B,CAAC;AAI7D,MAAM,WAAW,kBAAkB;IACjC,gBAAgB,EAAE,MAAM,CAAC;IACzB,GAAG,EAAE,MAAM,CAAC;CACb;AAED;;;GAGG;AACH,wBAAgB,YAAY,CAC1B,cAAc,EAAE,cAAc,EAC9B,OAAO,EAAE,aAAa,GACrB,kBAAkB,CAiCpB"}

package/dist/core/rag/capture-indexer.js

@@ -0,0 +1,37 @@
+/**
+ * Capture Indexer — indexes web capture results into the knowledge store.
+ * Called after captureWebPage() to persist extracted content for search.
+ */
+import { chunkText } from "./chunk-text.js";
+import { logger } from "../utils/logger.js";
+/**
+ * Index a web capture result into the knowledge store.
+ * Chunks the extracted text and stores with source_type='web_capture'.
+ */
+export function indexCapture(knowledgeStore, capture) {
+    const chunks = chunkText(capture.text);
+    if (chunks.length === 0) {
+        logger.info("No content to index from capture", { url: capture.url });
+        return { documentsIndexed: 0, url: capture.url };
+    }
+    const sourceId = `capture:${capture.url}`;
+    // Remove previous capture of this URL to avoid stale data
+    knowledgeStore.deleteBySource("web_capture", sourceId);
+    const docs = knowledgeStore.insertChunks(chunks.map((chunk) => ({
+        sourceType: "web_capture",
+        sourceId,
+        title: chunks.length > 1
+            ? `Web: ${capture.url} [${chunk.index + 1}/${chunks.length}]`
+            : `Web: ${capture.url}`,
+        content: chunk.content,
+        chunkIndex: chunk.index,
+        metadata: {
+            url: capture.url,
+            capturedAt: capture.capturedAt,
+            wordCount: capture.wordCount,
+        },
+    })));
+    logger.info("Web capture indexed", { url: capture.url, chunks: docs.length });
+    return { documentsIndexed: docs.length, url: capture.url };
+}
+//# sourceMappingURL=capture-indexer.js.map

package/dist/core/rag/capture-indexer.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"capture-indexer.js","sourceRoot":"","sources":["../../../src/core/rag/capture-indexer.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAIH,OAAO,EAAE,SAAS,EAAE,MAAM,iBAAiB,CAAC;AAC5C,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAO5C;;;GAGG;AACH,MAAM,UAAU,YAAY,CAC1B,cAA8B,EAC9B,OAAsB;IAEtB,MAAM,MAAM,GAAG,SAAS,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;IAEvC,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACxB,MAAM,CAAC,IAAI,CAAC,kCAAkC,EAAE,EAAE,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,CAAC,CAAC;QACtE,OAAO,EAAE,gBAAgB,EAAE,CAAC,EAAE,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,CAAC;IACnD,CAAC;IAED,MAAM,QAAQ,GAAG,WAAW,OAAO,CAAC,GAAG,EAAE,CAAC;IAE1C,0DAA0D;IAC1D,cAAc,CAAC,cAAc,CAAC,aAAa,EAAE,QAAQ,CAAC,CAAC;IAEvD,MAAM,IAAI,GAAG,cAAc,CAAC,YAAY,CACtC,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;QACrB,UAAU,EAAE,aAAsB;QAClC,QAAQ;QACR,KAAK,EAAE,MAAM,CAAC,MAAM,GAAG,CAAC;YACtB,CAAC,CAAC,QAAQ,OAAO,CAAC,GAAG,KAAK,KAAK,CAAC,KAAK,GAAG,CAAC,IAAI,MAAM,CAAC,MAAM,GAAG;YAC7D,CAAC,CAAC,QAAQ,OAAO,CAAC,GAAG,EAAE;QACzB,OAAO,EAAE,KAAK,CAAC,OAAO;QACtB,UAAU,EAAE,KAAK,CAAC,KAAK;QACvB,QAAQ,EAAE;YACR,GAAG,EAAE,OAAO,CAAC,GAAG;YAChB,UAAU,EAAE,OAAO,CAAC,UAAU;YAC9B,SAAS,EAAE,OAAO,CAAC,SAAS;SAC7B;KACF,CAAC,CAAC,CACJ,CAAC;IAEF,MAAM,CAAC,IAAI,CAAC,qBAAqB,EAAE,EAAE,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,MAAM,EAAE,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC;IAE9E,OAAO,EAAE,gBAAgB,EAAE,IAAI,CAAC,MAAM,EAAE,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,CAAC;AAC7D,CAAC"}

package/dist/core/rag/chunk-text.d.ts

@@ -0,0 +1,27 @@
+/**
+ * Text chunking utility — splits large documents into overlapping chunks
+ * for granular embedding and retrieval.
+ *
+ * Strategy: sentence-aware splitting with configurable chunk size and overlap.
+ * Default ~500 tokens/chunk (~2000 chars) with 10% overlap.
+ */
+export interface ChunkOptions {
+    /** Target tokens per chunk (default: 500) */
+    maxTokens?: number;
+    /** Overlap tokens between consecutive chunks (default: 50) */
+    overlapTokens?: number;
+}
+export interface TextChunk {
+    /** Zero-based chunk index */
+    index: number;
+    /** Chunk text content */
+    content: string;
+    /** Estimated token count */
+    tokens: number;
+}
+/**
+ * Split text into sentence-aware chunks with overlap.
+ * If text fits in a single chunk, returns one chunk.
+ */
+export declare function chunkText(text: string, options?: ChunkOptions): TextChunk[];
+//# sourceMappingURL=chunk-text.d.ts.map

package/dist/core/rag/chunk-text.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"chunk-text.d.ts","sourceRoot":"","sources":["../../../src/core/rag/chunk-text.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAIH,MAAM,WAAW,YAAY;IAC3B,6CAA6C;IAC7C,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,8DAA8D;IAC9D,aAAa,CAAC,EAAE,MAAM,CAAC;CACxB;AAED,MAAM,WAAW,SAAS;IACxB,6BAA6B;IAC7B,KAAK,EAAE,MAAM,CAAC;IACd,yBAAyB;IACzB,OAAO,EAAE,MAAM,CAAC;IAChB,4BAA4B;IAC5B,MAAM,EAAE,MAAM,CAAC;CAChB;AAKD;;;GAGG;AACH,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,YAAY,GAAG,SAAS,EAAE,CA+D3E"}

package/dist/core/rag/chunk-text.js

@@ -0,0 +1,112 @@
+/**
+ * Text chunking utility — splits large documents into overlapping chunks
+ * for granular embedding and retrieval.
+ *
+ * Strategy: sentence-aware splitting with configurable chunk size and overlap.
+ * Default ~500 tokens/chunk (~2000 chars) with 10% overlap.
+ */
+import { estimateTokens } from "../context/token-estimator.js";
+const DEFAULT_MAX_TOKENS = 500;
+const DEFAULT_OVERLAP_TOKENS = 50;
+/**
+ * Split text into sentence-aware chunks with overlap.
+ * If text fits in a single chunk, returns one chunk.
+ */
+export function chunkText(text, options) {
+    const maxTokens = options?.maxTokens ?? DEFAULT_MAX_TOKENS;
+    const overlapTokens = options?.overlapTokens ?? DEFAULT_OVERLAP_TOKENS;
+    const trimmed = text.trim();
+    if (!trimmed)
+        return [];
+    const totalTokens = estimateTokens(trimmed);
+    if (totalTokens <= maxTokens) {
+        return [{ index: 0, content: trimmed, tokens: totalTokens }];
+    }
+    // Split into sentences (preserve the delimiter)
+    const sentences = splitSentences(trimmed);
+    const chunks = [];
+    let currentSentences = [];
+    let currentTokens = 0;
+    for (const sentence of sentences) {
+        const sentTokens = estimateTokens(sentence);
+        // If a single sentence exceeds maxTokens, split it by chars
+        if (sentTokens > maxTokens && currentSentences.length === 0) {
+            const hardChunks = hardSplit(sentence, maxTokens, overlapTokens);
+            for (const hc of hardChunks) {
+                chunks.push({ index: chunks.length, content: hc, tokens: estimateTokens(hc) });
+            }
+            continue;
+        }
+        if (currentTokens + sentTokens > maxTokens && currentSentences.length > 0) {
+            // Emit current chunk
+            const chunkContent = currentSentences.join("").trim();
+            chunks.push({
+                index: chunks.length,
+                content: chunkContent,
+                tokens: estimateTokens(chunkContent),
+            });
+            // Build overlap from tail of current sentences
+            const overlapResult = buildOverlap(currentSentences, overlapTokens);
+            currentSentences = overlapResult;
+            currentTokens = estimateTokens(overlapResult.join(""));
+        }
+        currentSentences.push(sentence);
+        currentTokens += sentTokens;
+    }
+    // Emit last chunk
+    if (currentSentences.length > 0) {
+        const chunkContent = currentSentences.join("").trim();
+        if (chunkContent) {
+            chunks.push({
+                index: chunks.length,
+                content: chunkContent,
+                tokens: estimateTokens(chunkContent),
+            });
+        }
+    }
+    return chunks;
+}
+/**
+ * Split text into sentences. Keeps delimiters attached to the sentence.
+ */
+function splitSentences(text) {
+    // Split on sentence boundaries but keep the delimiter
+    const parts = text.split(/(?<=[.!?\n])\s+/);
+    return parts.filter((p) => p.length > 0);
+}
+/**
+ * Hard-split a long string into chunks of roughly maxTokens.
+ */
+function hardSplit(text, maxTokens, overlapTokens) {
+    const maxChars = maxTokens * 4; // ~4 chars per token
+    const overlapChars = overlapTokens * 4;
+    const chunks = [];
+    let start = 0;
+    while (start < text.length) {
+        const end = Math.min(start + maxChars, text.length);
+        chunks.push(text.slice(start, end));
+        start = end - overlapChars;
+        if (start >= text.length)
+            break;
+        // Prevent infinite loop if overlap >= chunk size
+        if (end === text.length)
+            break;
+    }
+    return chunks;
+}
+/**
+ * Build overlap sentences from the tail of the current sentence list.
+ */
+function buildOverlap(sentences, overlapTokens) {
+    const result = [];
+    let tokens = 0;
+    for (let i = sentences.length - 1; i >= 0; i--) {
+        const sentTokens = estimateTokens(sentences[i]);
+        if (tokens + sentTokens > overlapTokens && result.length > 0)
+            break;
+        result.unshift(sentences[i]);
+        tokens += sentTokens;
+    }
+    return result;
+}
+//# sourceMappingURL=chunk-text.js.map

package/dist/core/rag/chunk-text.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"chunk-text.js","sourceRoot":"","sources":["../../../src/core/rag/chunk-text.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH,OAAO,EAAE,cAAc,EAAE,MAAM,+BAA+B,CAAC;AAkB/D,MAAM,kBAAkB,GAAG,GAAG,CAAC;AAC/B,MAAM,sBAAsB,GAAG,EAAE,CAAC;AAElC;;;GAGG;AACH,MAAM,UAAU,SAAS,CAAC,IAAY,EAAE,OAAsB;IAC5D,MAAM,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,kBAAkB,CAAC;IAC3D,MAAM,aAAa,GAAG,OAAO,EAAE,aAAa,IAAI,sBAAsB,CAAC;IAEvE,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;IAC5B,IAAI,CAAC,OAAO;QAAE,OAAO,EAAE,CAAC;IAExB,MAAM,WAAW,GAAG,cAAc,CAAC,OAAO,CAAC,CAAC;IAC5C,IAAI,WAAW,IAAI,SAAS,EAAE,CAAC;QAC7B,OAAO,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,WAAW,EAAE,CAAC,CAAC;IAC/D,CAAC;IAED,gDAAgD;IAChD,MAAM,SAAS,GAAG,cAAc,CAAC,OAAO,CAAC,CAAC;IAE1C,MAAM,MAAM,GAAgB,EAAE,CAAC;IAC/B,IAAI,gBAAgB,GAAa,EAAE,CAAC;IACpC,IAAI,aAAa,GAAG,CAAC,CAAC;IAEtB,KAAK,MAAM,QAAQ,IAAI,SAAS,EAAE,CAAC;QACjC,MAAM,UAAU,GAAG,cAAc,CAAC,QAAQ,CAAC,CAAC;QAE5C,4DAA4D;QAC5D,IAAI,UAAU,GAAG,SAAS,IAAI,gBAAgB,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YAC5D,MAAM,UAAU,GAAG,SAAS,CAAC,QAAQ,EAAE,SAAS,EAAE,aAAa,CAAC,CAAC;YACjE,KAAK,MAAM,EAAE,IAAI,UAAU,EAAE,CAAC;gBAC5B,MAAM,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,EAAE,EAAE,EAAE,MAAM,EAAE,cAAc,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;YACjF,CAAC;YACD,SAAS;QACX,CAAC;QAED,IAAI,aAAa,GAAG,UAAU,GAAG,SAAS,IAAI,gBAAgB,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC1E,qBAAqB;YACrB,MAAM,YAAY,GAAG,gBAAgB,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;YACtD,MAAM,CAAC,IAAI,CAAC;gBACV,KAAK,EAAE,MAAM,CAAC,MAAM;gBACpB,OAAO,EAAE,YAAY;gBACrB,MAAM,EAAE,cAAc,CAAC,YAAY,CAAC;aACrC,CAAC,CAAC;YAEH,+CAA+C;YAC/C,MAAM,aAAa,GAAG,YAAY,CAAC,gBAAgB,EAAE,aAAa,CAAC,CAAC;YACpE,gBAAgB,GAAG,aAAa,CAAC;YACjC,aAAa,GAAG,cAAc,CAAC,aAAa,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC;QACzD,CAAC;QAED,gBAAgB,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;QAChC,aAAa,IAAI,UAAU,CAAC;IAC9B,CAAC;IAED,kBAAkB;IAClB,IAAI,gBAAgB,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAChC,MAAM,YAAY,GAAG,gBAAgB,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;QACtD,IAAI,YAAY,EAAE,CAAC;YACjB,MAAM,CAAC,IAAI,CAAC;gBACV,KAAK,EAAE,MAAM,CAAC,MAAM;gBACpB,OAAO,EAAE,YAAY;gBACrB,MAAM,EAAE,cAAc,CAAC,YAAY,CAAC;aACrC,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAED;;GAEG;AACH,SAAS,cAAc,CAAC,IAAY;IAClC,sDAAsD;IACtD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,iBAAiB,CAAC,CAAC;IAC5C,OAAO,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;AAC3C,CAAC;AAED;;GAEG;AACH,SAAS,SAAS,CAAC,IAAY,EAAE,SAAiB,EAAE,aAAqB;IACvE,MAAM,QAAQ,GAAG,SAAS,GAAG,CAAC,CAAC,CAAC,qBAAqB;IACrD,MAAM,YAAY,GAAG,aAAa,GAAG,CAAC,CAAC;IACvC,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,OAAO,KAAK,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QAC3B,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,QAAQ,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC;QACpD,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,CAAC;QACpC,KAAK,GAAG,GAAG,GAAG,YAAY,CAAC;QAC3B,IAAI,KAAK,IAAI,IAAI,CAAC,MAAM;YAAE,MAAM;QAChC,iDAAiD;QACjD,IAAI,GAAG,KAAK,IAAI,CAAC,MAAM;YAAE,MAAM;IACjC,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAED;;GAEG;AACH,SAAS,YAAY,CAAC,SAAmB,EAAE,aAAqB;IAC9D,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,IAAI,MAAM,GAAG,CAAC,CAAC;IAEf,KAAK,IAAI,CAAC,GAAG,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC/C,MAAM,UAAU,GAAG,cAAc,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC;QAChD,IAAI,MAAM,GAAG,UAAU,GAAG,aAAa,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;YAAE,MAAM;QACpE,MAAM,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC;QAC7B,MAAM,IAAI,UAAU,CAAC;IACvB,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC"}

package/dist/core/rag/docs-indexer.d.ts

@@ -0,0 +1,18 @@
+/**
+ * Docs Indexer — mirrors docs_cache entries into the knowledge store.
+ * Keeps backward compatibility with DocsCacheStore while enabling
+ * unified search across all knowledge sources.
+ */
+import { DocsCacheStore } from "../docs/docs-cache-store.js";
+import { KnowledgeStore } from "../store/knowledge-store.js";
+export interface DocsIndexResult {
+    docsFound: number;
+    documentsIndexed: number;
+    skippedDuplicates: number;
+}
+/**
+ * Index all cached docs into the knowledge store.
+ * Each cached doc is chunked and stored with source_type='docs'.
+ */
+export declare function indexCachedDocs(knowledgeStore: KnowledgeStore, docsCacheStore: DocsCacheStore): DocsIndexResult;
+//# sourceMappingURL=docs-indexer.d.ts.map

package/dist/core/rag/docs-indexer.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"docs-indexer.d.ts","sourceRoot":"","sources":["../../../src/core/rag/docs-indexer.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,EAAE,cAAc,EAAE,MAAM,6BAA6B,CAAC;AAC7D,OAAO,EAAE,cAAc,EAAE,MAAM,6BAA6B,CAAC;AAI7D,MAAM,WAAW,eAAe;IAC9B,SAAS,EAAE,MAAM,CAAC;IAClB,gBAAgB,EAAE,MAAM,CAAC;IACzB,iBAAiB,EAAE,MAAM,CAAC;CAC3B;AAED;;;GAGG;AACH,wBAAgB,eAAe,CAC7B,cAAc,EAAE,cAAc,EAC9B,cAAc,EAAE,cAAc,GAC7B,eAAe,CAoDjB"}

package/dist/core/rag/docs-indexer.js

@@ -0,0 +1,57 @@
+/**
+ * Docs Indexer — mirrors docs_cache entries into the knowledge store.
+ * Keeps backward compatibility with DocsCacheStore while enabling
+ * unified search across all knowledge sources.
+ */
+import { chunkText } from "./chunk-text.js";
+import { logger } from "../utils/logger.js";
+/**
+ * Index all cached docs into the knowledge store.
+ * Each cached doc is chunked and stored with source_type='docs'.
+ */
+export function indexCachedDocs(knowledgeStore, docsCacheStore) {
+    const allDocs = docsCacheStore.listCached();
+    if (allDocs.length === 0) {
+        logger.info("No cached docs to index");
+        return { docsFound: 0, documentsIndexed: 0, skippedDuplicates: 0 };
+    }
+    let documentsIndexed = 0;
+    let skippedDuplicates = 0;
+    for (const doc of allDocs) {
+        const sourceId = `docs:${doc.libId}`;
+        const chunks = chunkText(doc.content);
+        // Remove previous version to re-index fresh content
+        knowledgeStore.deleteBySource("docs", sourceId);
+        for (const chunk of chunks) {
+            const countBefore = knowledgeStore.count("docs");
+            knowledgeStore.insert({
+                sourceType: "docs",
+                sourceId,
+                title: chunks.length > 1
+                    ? `${doc.libName} [${chunk.index + 1}/${chunks.length}]`
+                    : doc.libName,
+                content: chunk.content,
+                chunkIndex: chunk.index,
+                metadata: {
+                    libId: doc.libId,
+                    libName: doc.libName,
+                    version: doc.version,
+                    fetchedAt: doc.fetchedAt,
+                },
+            });
+            if (knowledgeStore.count("docs") > countBefore) {
+                documentsIndexed++;
+            }
+            else {
+                skippedDuplicates++;
+            }
+        }
+    }
+    logger.info("Cached docs indexed", {
+        docsFound: allDocs.length,
+        documentsIndexed,
+        skippedDuplicates,
+    });
+    return { docsFound: allDocs.length, documentsIndexed, skippedDuplicates };
+}
+//# sourceMappingURL=docs-indexer.js.map

package/dist/core/rag/docs-indexer.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"docs-indexer.js","sourceRoot":"","sources":["../../../src/core/rag/docs-indexer.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAIH,OAAO,EAAE,SAAS,EAAE,MAAM,iBAAiB,CAAC;AAC5C,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAQ5C;;;GAGG;AACH,MAAM,UAAU,eAAe,CAC7B,cAA8B,EAC9B,cAA8B;IAE9B,MAAM,OAAO,GAAG,cAAc,CAAC,UAAU,EAAE,CAAC;IAE5C,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACzB,MAAM,CAAC,IAAI,CAAC,yBAAyB,CAAC,CAAC;QACvC,OAAO,EAAE,SAAS,EAAE,CAAC,EAAE,gBAAgB,EAAE,CAAC,EAAE,iBAAiB,EAAE,CAAC,EAAE,CAAC;IACrE,CAAC;IAED,IAAI,gBAAgB,GAAG,CAAC,CAAC;IACzB,IAAI,iBAAiB,GAAG,CAAC,CAAC;IAE1B,KAAK,MAAM,GAAG,IAAI,OAAO,EAAE,CAAC;QAC1B,MAAM,QAAQ,GAAG,QAAQ,GAAG,CAAC,KAAK,EAAE,CAAC;QACrC,MAAM,MAAM,GAAG,SAAS,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;QAEtC,oDAAoD;QACpD,cAAc,CAAC,cAAc,CAAC,MAAM,EAAE,QAAQ,CAAC,CAAC;QAEhD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;YAC3B,MAAM,WAAW,GAAG,cAAc,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC;YAEjD,cAAc,CAAC,MAAM,CAAC;gBACpB,UAAU,EAAE,MAAM;gBAClB,QAAQ;gBACR,KAAK,EAAE,MAAM,CAAC,MAAM,GAAG,CAAC;oBACtB,CAAC,CAAC,GAAG,GAAG,CAAC,OAAO,KAAK,KAAK,CAAC,KAAK,GAAG,CAAC,IAAI,MAAM,CAAC,MAAM,GAAG;oBACxD,CAAC,CAAC,GAAG,CAAC,OAAO;gBACf,OAAO,EAAE,KAAK,CAAC,OAAO;gBACtB,UAAU,EAAE,KAAK,CAAC,KAAK;gBACvB,QAAQ,EAAE;oBACR,KAAK,EAAE,GAAG,CAAC,KAAK;oBAChB,OAAO,EAAE,GAAG,CAAC,OAAO;oBACpB,OAAO,EAAE,GAAG,CAAC,OAAO;oBACpB,SAAS,EAAE,GAAG,CAAC,SAAS;iBACzB;aACF,CAAC,CAAC;YAEH,IAAI,cAAc,CAAC,KAAK,CAAC,MAAM,CAAC,GAAG,WAAW,EAAE,CAAC;gBAC/C,gBAAgB,EAAE,CAAC;YACrB,CAAC;iBAAM,CAAC;gBACN,iBAAiB,EAAE,CAAC;YACtB,CAAC;QACH,CAAC;IACH,CAAC;IAED,MAAM,CAAC,IAAI,CAAC,qBAAqB,EAAE;QACjC,SAAS,EAAE,OAAO,CAAC,MAAM;QACzB,gBAAgB;QAChB,iBAAiB;KAClB,CAAC,CAAC;IAEH,OAAO,EAAE,SAAS,EAAE,OAAO,CAAC,MAAM,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,CAAC;AAC5E,CAAC"}

package/dist/core/rag/embedding-store.d.ts

@@ -0,0 +1,51 @@
+/**
+ * Embedding store — persists vector embeddings in SQLite.
+ * Supports cosine similarity search for semantic retrieval.
+ * All data stays local — no external API calls.
+ */
+import type { SqliteStore } from "../store/sqlite-store.js";
+export interface EmbeddingEntry {
+    id: string;
+    source: string;
+    sourceId: string;
+    text: string;
+    embedding: number[];
+}
+export interface SimilarityResult {
+    id: string;
+    source: string;
+    sourceId: string;
+    text: string;
+    similarity: number;
+}
+export declare class EmbeddingStore {
+    private sqliteStore;
+    private db;
+    constructor(sqliteStore: SqliteStore);
+    /**
+     * Insert or update an embedding entry.
+     */
+    upsert(entry: EmbeddingEntry): void;
+    /**
+     * Get an embedding entry by ID.
+     */
+    getById(id: string): EmbeddingEntry | null;
+    /**
+     * Find the most similar embeddings to a query vector using cosine similarity.
+     * Performs brute-force search (suitable for < 100k embeddings).
+     */
+    findSimilar(queryVector: number[], limit?: number): SimilarityResult[];
+    /**
+     * Count stored embeddings.
+     */
+    count(): number;
+    /**
+     * Delete an embedding by ID.
+     */
+    delete(id: string): void;
+    /**
+     * Clear all embeddings.
+     */
+    clear(): void;
+}
+//# sourceMappingURL=embedding-store.d.ts.map

package/dist/core/rag/embedding-store.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"embedding-store.d.ts","sourceRoot":"","sources":["../../../src/core/rag/embedding-store.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAGH,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,0BAA0B,CAAC;AAG5D,MAAM,WAAW,cAAc;IAC7B,EAAE,EAAE,MAAM,CAAC;IACX,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,MAAM,CAAC;IACb,SAAS,EAAE,MAAM,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,gBAAgB;IAC/B,EAAE,EAAE,MAAM,CAAC;IACX,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,CAAC;IACjB,IAAI,EAAE,MAAM,CAAC;IACb,UAAU,EAAE,MAAM,CAAC;CACpB;AAiED,qBAAa,cAAc;IAGb,OAAO,CAAC,WAAW;IAF/B,OAAO,CAAC,EAAE,CAAoB;gBAEV,WAAW,EAAE,WAAW;IAK5C;;OAEG;IACH,MAAM,CAAC,KAAK,EAAE,cAAc,GAAG,IAAI;IAmBnC;;OAEG;IACH,OAAO,CAAC,EAAE,EAAE,MAAM,GAAG,cAAc,GAAG,IAAI;IAiB1C;;;OAGG;IACH,WAAW,CAAC,WAAW,EAAE,MAAM,EAAE,EAAE,KAAK,GAAE,MAAW,GAAG,gBAAgB,EAAE;IAoB1E;;OAEG;IACH,KAAK,IAAI,MAAM;IAQf;;OAEG;IACH,MAAM,CAAC,EAAE,EAAE,MAAM,GAAG,IAAI;IAOxB;;OAEG;IACH,KAAK,IAAI,IAAI;CAGd"}

package/dist/core/rag/embedding-store.js

@@ -0,0 +1,134 @@
+/**
+ * Embedding store — persists vector embeddings in SQLite.
+ * Supports cosine similarity search for semantic retrieval.
+ * All data stays local — no external API calls.
+ */
+const CREATE_TABLE_SQL = `
+  CREATE TABLE IF NOT EXISTS embeddings (
+    id TEXT PRIMARY KEY,
+    source TEXT NOT NULL,
+    source_id TEXT NOT NULL,
+    text TEXT NOT NULL,
+    embedding BLOB NOT NULL,
+    created_at TEXT NOT NULL DEFAULT (datetime('now'))
+  )
+`;
+/**
+ * Compute cosine similarity between two vectors.
+ */
+function cosineSimilarity(a, b) {
+    if (a.length !== b.length || a.length === 0)
+        return 0;
+    let dotProduct = 0;
+    let normA = 0;
+    let normB = 0;
+    for (let i = 0; i < a.length; i++) {
+        dotProduct += a[i] * b[i];
+        normA += a[i] * a[i];
+        normB += b[i] * b[i];
+    }
+    const denominator = Math.sqrt(normA) * Math.sqrt(normB);
+    if (denominator === 0)
+        return 0;
+    return dotProduct / denominator;
+}
+/**
+ * Serialize a number array to a Buffer for SQLite BLOB storage.
+ */
+function serializeVector(vec) {
+    const buf = Buffer.alloc(vec.length * 8);
+    for (let i = 0; i < vec.length; i++) {
+        buf.writeDoubleLE(vec[i], i * 8);
+    }
+    return buf;
+}
+/**
+ * Deserialize a Buffer from SQLite BLOB to a number array.
+ */
+function deserializeVector(buf) {
+    const vec = [];
+    for (let i = 0; i < buf.length; i += 8) {
+        vec.push(buf.readDoubleLE(i));
+    }
+    return vec;
+}
+export class EmbeddingStore {
+    sqliteStore;
+    db;
+    constructor(sqliteStore) {
+        this.sqliteStore = sqliteStore;
+        this.db = this.sqliteStore.getDb();
+        this.db.exec(CREATE_TABLE_SQL);
+    }
+    /**
+     * Insert or update an embedding entry.
+     */
+    upsert(entry) {
+        const stmt = this.db.prepare(`
+      INSERT INTO embeddings (id, source, source_id, text, embedding)
+      VALUES (?, ?, ?, ?, ?)
+      ON CONFLICT(id) DO UPDATE SET
+        source = excluded.source,
+        source_id = excluded.source_id,
+        text = excluded.text,
+        embedding = excluded.embedding
+    `);
+        stmt.run(entry.id, entry.source, entry.sourceId, entry.text, serializeVector(entry.embedding));
+    }
+    /**
+     * Get an embedding entry by ID.
+     */
+    getById(id) {
+        const stmt = this.db.prepare("SELECT id, source, source_id, text, embedding FROM embeddings WHERE id = ?");
+        const row = stmt.get(id);
+        if (!row)
+            return null;
+        return {
+            id: row.id,
+            source: row.source,
+            sourceId: row.source_id,
+            text: row.text,
+            embedding: deserializeVector(row.embedding),
+        };
+    }
+    /**
+     * Find the most similar embeddings to a query vector using cosine similarity.
+     * Performs brute-force search (suitable for < 100k embeddings).
+     */
+    findSimilar(queryVector, limit = 10) {
+        const stmt = this.db.prepare("SELECT id, source, source_id, text, embedding FROM embeddings");
+        const rows = stmt.all();
+        const results = rows.map((row) => ({
+            id: row.id,
+            source: row.source,
+            sourceId: row.source_id,
+            text: row.text,
+            similarity: cosineSimilarity(queryVector, deserializeVector(row.embedding)),
+        }));
+        // Sort by similarity descending
+        results.sort((a, b) => b.similarity - a.similarity);
+        return results.slice(0, limit);
+    }
+    /**
+     * Count stored embeddings.
+     */
+    count() {
+        const stmt = this.db.prepare("SELECT COUNT(*) as cnt FROM embeddings");
+        const row = stmt.get();
+        return row.cnt;
+    }
+    /**
+     * Delete an embedding by ID.
+     */
+    delete(id) {
+        const stmt = this.db.prepare("DELETE FROM embeddings WHERE id = ?");
+        stmt.run(id);
+    }
+    /**
+     * Clear all embeddings.
+     */
+    clear() {
+        this.db.exec("DELETE FROM embeddings");
+    }
+}
+//# sourceMappingURL=embedding-store.js.map

package/dist/core/rag/embedding-store.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"embedding-store.js","sourceRoot":"","sources":["../../../src/core/rag/embedding-store.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAsBH,MAAM,gBAAgB,GAAG;;;;;;;;;CASxB,CAAC;AAEF;;GAEG;AACH,SAAS,gBAAgB,CAAC,CAAW,EAAE,CAAW;IAChD,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IAEtD,IAAI,UAAU,GAAG,CAAC,CAAC;IACnB,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAClC,UAAU,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QAC1B,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QACrB,KAAK,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IACvB,CAAC;IAED,MAAM,WAAW,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACxD,IAAI,WAAW,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IAEhC,OAAO,UAAU,GAAG,WAAW,CAAC;AAClC,CAAC;AAED;;GAEG;AACH,SAAS,eAAe,CAAC,GAAa;IACpC,MAAM,GAAG,GAAG,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACzC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACpC,GAAG,CAAC,aAAa,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC;IACnC,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED;;GAEG;AACH,SAAS,iBAAiB,CAAC,GAAW;IACpC,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,CAAC,MAAM,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;QACvC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC;IAChC,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAUD,MAAM,OAAO,cAAc;IAGL;IAFZ,EAAE,CAAoB;IAE9B,YAAoB,WAAwB;QAAxB,gBAAW,GAAX,WAAW,CAAa;QAC1C,IAAI,CAAC,EAAE,GAAG,IAAI,CAAC,WAAW,CAAC,KAAK,EAAE,CAAC;QACnC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,gBAAgB,CAAC,CAAC;IACjC,CAAC;IAED;;OAEG;IACH,MAAM,CAAC,KAAqB;QAC1B,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC;;;;;;;;KAQ5B,CAAC,CAAC;QACH,IAAI,CAAC,GAAG,CACN,KAAK,CAAC,EAAE,EACR,KAAK,CAAC,MAAM,EACZ,KAAK,CAAC,QAAQ,EACd,KAAK,CAAC,IAAI,EACV,eAAe,CAAC,KAAK,CAAC,SAAS,CAAC,CACjC,CAAC;IACJ,CAAC;IAED;;OAEG;IACH,OAAO,CAAC,EAAU;QAChB,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,4EAA4E,CAC7E,CAAC;QACF,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,EAAE,CAA6B,CAAC;QAErD,IAAI,CAAC,GAAG;YAAE,OAAO,IAAI,CAAC;QAEtB,OAAO;YACL,EAAE,EAAE,GAAG,CAAC,EAAE;YACV,MAAM,EAAE,GAAG,CAAC,MAAM;YAClB,QAAQ,EAAE,GAAG,CAAC,SAAS;YACvB,IAAI,EAAE,GAAG,CAAC,IAAI;YACd,SAAS,EAAE,iBAAiB,CAAC,GAAG,CAAC,SAAS,CAAC;SAC5C,CAAC;IACJ,CAAC;IAED;;;OAGG;IACH,WAAW,CAAC,WAAqB,EAAE,QAAgB,EAAE;QACnD,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,+DAA+D,CAChE,CAAC;QACF,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,EAAoB,CAAC;QAE1C,MAAM,OAAO,GAAuB,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC;YACrD,EAAE,EAAE,GAAG,CAAC,EAAE;YACV,MAAM,EAAE,GAAG,CAAC,MAAM;YAClB,QAAQ,EAAE,GAAG,CAAC,SAAS;YACvB,IAAI,EAAE,GAAG,CAAC,IAAI;YACd,UAAU,EAAE,gBAAgB,CAAC,WAAW,EAAE,iBAAiB,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;SAC5E,CAAC,CAAC,CAAC;QAEJ,gCAAgC;QAChC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,UAAU,GAAG,CAAC,CAAC,UAAU,CAAC,CAAC;QAEpD,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;IACjC,CAAC;IAED;;OAEG;IACH,KAAK;QACH,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,wCAAwC,CACzC,CAAC;QACF,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,EAAqB,CAAC;QAC1C,OAAO,GAAG,CAAC,GAAG,CAAC;IACjB,CAAC;IAED;;OAEG;IACH,MAAM,CAAC,EAAU;QACf,MAAM,IAAI,GAAG,IAAI,CAAC,EAAE,CAAC,OAAO,CAC1B,qCAAqC,CACtC,CAAC;QACF,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;IACf,CAAC;IAED;;OAEG;IACH,KAAK;QACH,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,wBAAwB,CAAC,CAAC;IACzC,CAAC;CACF"}

package/dist/core/rag/rag-pipeline.d.ts

@@ -0,0 +1,48 @@
+/**
+ * RAG Semantic Pipeline — indexes nodes as embeddings and provides semantic search.
+ *
+ * Strategy: Uses lightweight TF-IDF vectorization for embeddings (no external deps).
+ * This provides semantic-like search without requiring transformers.js (~400MB).
+ * Can be upgraded to real transformer embeddings later.
+ *
+ * All processing is local — no external API calls.
+ */
+import type { SqliteStore } from "../store/sqlite-store.js";
+import { EmbeddingStore, type SimilarityResult } from "./embedding-store.js";
+/**
+ * TF-IDF Vectorizer — encapsulates vocabulary and IDF state per instance.
+ * Avoids module-level mutable state.
+ */
+export declare class TfIdfVectorizer {
+    private vocab;
+    private idf;
+    get vocabSize(): number;
+    /**
+     * Build vocabulary and IDF from tokenized documents.
+     */
+    fit(documents: string[][]): void;
+    /**
+     * Generate an embedding for text using TF-IDF vectorization.
+     * Falls back to hash-based embedding if vocabulary not built.
+     */
+    embed(text: string): number[];
+}
+/**
+ * Index all nodes from the store as embeddings.
+ * Builds a TF-IDF vocabulary from all node texts, then stores vectors.
+ */
+export declare function indexNodeEmbeddings(store: SqliteStore, embeddingStore: EmbeddingStore): Promise<number>;
+/**
+ * Index all sources (nodes + knowledge documents) as embeddings with a unified vocabulary.
+ * This produces cross-source similarity by fitting TF-IDF on the combined corpus.
+ */
+export declare function indexAllEmbeddings(store: SqliteStore, embeddingStore: EmbeddingStore): Promise<{
+    nodes: number;
+    knowledge: number;
+}>;
+/**
+ * Semantic search using embeddings.
+ * Embeds the query and finds similar documents.
+ */
+export declare function semanticSearch(embeddingStore: EmbeddingStore, query: string, limit?: number): Promise<SimilarityResult[]>;
+//# sourceMappingURL=rag-pipeline.d.ts.map

package/dist/core/rag/rag-pipeline.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"rag-pipeline.d.ts","sourceRoot":"","sources":["../../../src/core/rag/rag-pipeline.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,0BAA0B,CAAC;AAE5D,OAAO,EAAE,cAAc,EAAE,KAAK,gBAAgB,EAAE,MAAM,sBAAsB,CAAC;AAgG7E;;;GAGG;AACH,qBAAa,eAAe;IAC1B,OAAO,CAAC,KAAK,CAAoC;IACjD,OAAO,CAAC,GAAG,CAAoC;IAE/C,IAAI,SAAS,IAAI,MAAM,CAEtB;IAED;;OAEG;IACH,GAAG,CAAC,SAAS,EAAE,MAAM,EAAE,EAAE,GAAG,IAAI;IAMhC;;;OAGG;IACH,KAAK,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE;CAM9B;AAsCD;;;GAGG;AACH,wBAAsB,mBAAmB,CACvC,KAAK,EAAE,WAAW,EAClB,cAAc,EAAE,cAAc,GAC7B,OAAO,CAAC,MAAM,CAAC,CAgDjB;AAED;;;GAGG;AACH,wBAAsB,kBAAkB,CACtC,KAAK,EAAE,WAAW,EAClB,cAAc,EAAE,cAAc,GAC7B,OAAO,CAAC;IAAE,KAAK,EAAE,MAAM,CAAC;IAAC,SAAS,EAAE,MAAM,CAAA;CAAE,CAAC,CAwE/C;AAED;;;GAGG;AACH,wBAAsB,cAAc,CAClC,cAAc,EAAE,cAAc,EAC9B,KAAK,EAAE,MAAM,EACb,KAAK,GAAE,MAAW,GACjB,OAAO,CAAC,gBAAgB,EAAE,CAAC,CAS7B"}