@framers/agentos 0.6.30 → 0.7.0

package/dist/cognition/memory/io/ingestion/IDocumentLoader.d.ts

@@ -0,0 +1,63 @@
+/**
+ * @fileoverview IDocumentLoader — contract for all document loading strategies.
+ *
+ * Every loader targets one or more file extensions and exposes a uniform
+ * `load()` primitive that normalises diverse document formats into a single
+ * {@link LoadedDocument} shape ready for chunking and indexing.
+ *
+ * @module memory/ingestion/IDocumentLoader
+ */
+import type { LoadOptions, LoadedDocument } from '../facade/types.js';
+/**
+ * Unified interface for all document loaders in the AgentOS ingestion
+ * pipeline.
+ *
+ * Implementations handle a specific set of file extensions and are
+ * responsible for:
+ * 1. Reading raw bytes from a file path or in-memory `Buffer`.
+ * 2. Extracting plain-text `content` and structured `metadata`.
+ * 3. Returning a {@link LoadedDocument} ready for downstream chunking.
+ *
+ * @example
+ * ```ts
+ * const loader: IDocumentLoader = new MarkdownLoader();
+ * if (loader.canLoad('README.md')) {
+ *   const doc = await loader.load('README.md');
+ *   console.log(doc.metadata.title);
+ * }
+ * ```
+ */
+export interface IDocumentLoader {
+    /**
+     * File extensions this loader handles, each with a leading dot.
+     *
+     * Used by {@link LoaderRegistry} to route file paths to the correct loader.
+     *
+     * @example ['.md', '.mdx']
+     */
+    readonly supportedExtensions: string[];
+    /**
+     * Parses `source` and returns a normalised {@link LoadedDocument}.
+     *
+     * When `source` is a `string` the loader treats it as an absolute (or
+     * resolvable) file path and reads the file from disk.  When `source` is a
+     * `Buffer` the loader parses the bytes directly and derives as much
+     * metadata as possible from the buffer content alone.
+     *
+     * @param source  - Absolute file path OR raw document bytes.
+     * @param options - Optional hints such as a format override.
+     * @returns A promise resolving to the fully-populated {@link LoadedDocument}.
+     * @throws When the file cannot be read or the format is not parsable.
+     */
+    load(source: string | Buffer, options?: LoadOptions): Promise<LoadedDocument>;
+    /**
+     * Returns `true` when this loader is capable of handling `source`.
+     *
+     * For string sources the check is purely extension-based.  For `Buffer`
+     * sources the loader may inspect magic bytes when relevant.
+     *
+     * @param source - Absolute file path or raw bytes.
+     */
+    canLoad(source: string | Buffer): boolean;
+}
+//# sourceMappingURL=IDocumentLoader.d.ts.map

package/dist/cognition/memory/io/ingestion/IDocumentLoader.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"IDocumentLoader.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/IDocumentLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,KAAK,EAAE,WAAW,EAAE,cAAc,EAAE,MAAM,oBAAoB,CAAC;AAEtE;;;;;;;;;;;;;;;;;;GAkBG;AACH,MAAM,WAAW,eAAe;IAC9B;;;;;;OAMG;IACH,QAAQ,CAAC,mBAAmB,EAAE,MAAM,EAAE,CAAC;IAEvC;;;;;;;;;;;;OAYG;IACH,IAAI,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,OAAO,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,cAAc,CAAC,CAAC;IAE9E;;;;;;;OAOG;IACH,OAAO,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,GAAG,OAAO,CAAC;CAC3C"}

package/dist/cognition/memory/io/ingestion/IDocumentLoader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"IDocumentLoader.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/IDocumentLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG"}

package/dist/cognition/memory/io/ingestion/LoaderRegistry.d.ts

@@ -0,0 +1,140 @@
+/**
+ * @fileoverview LoaderRegistry — extension-to-loader routing table.
+ *
+ * The registry maintains a map of file extensions to {@link IDocumentLoader}
+ * implementations and provides a convenience `loadFile()` method that
+ * auto-detects the format from a file path before delegating to the
+ * appropriate loader.
+ *
+ * On construction the registry pre-registers five built-in loaders:
+ * {@link TextLoader}, {@link MarkdownLoader}, {@link HtmlLoader},
+ * {@link PdfLoader}, and {@link DocxLoader}.  In addition, the optional
+ * OCR and Docling-backed loaders are registered when their
+ * respective factories return non-null values (i.e. when `tesseract.js` and
+ * `python3 -m docling` are available in the environment).
+ *
+ * Additional loaders can be added at runtime via {@link LoaderRegistry.register}.
+ *
+ * @module memory/ingestion/LoaderRegistry
+ */
+import type { IDocumentLoader } from './IDocumentLoader.js';
+import type { LoadOptions, LoadedDocument } from '../facade/types.js';
+/**
+ * Central registry mapping file extensions to {@link IDocumentLoader}
+ * implementations.
+ *
+ * ### Built-in loaders (registered automatically)
+ * | Extensions                                         | Loader                |
+ * |----------------------------------------------------|-----------------------|
+ * | `.txt`, `.csv`, `.tsv`, `.json`, `.yaml`, `.yml`  | {@link TextLoader}    |
+ * | `.md`, `.mdx`                                     | {@link MarkdownLoader} |
+ * | `.html`, `.htm`                                   | {@link HtmlLoader}    |
+ * | `.pdf`                                            | {@link PdfLoader}     |
+ * | `.docx`                                           | {@link DocxLoader}    |
+ *
+ * ### Conditional loaders (registered when available)
+ * | Condition                     | Loader                                                  |
+ * |-------------------------------|---------------------------------------------------------|
+ * | `tesseract.js` installed      | factory from {@link createOcrPdfLoader} (overrides PDF) |
+ * | `python3 -m docling` available | factory from {@link createDoclingLoader} (overrides PDF + DOCX) |
+ *
+ * ### Registering a custom loader
+ * ```ts
+ * const registry = new LoaderRegistry();
+ * registry.register(new PdfLoader());
+ * const doc = await registry.loadFile('/reports/q3.pdf');
+ * ```
+ *
+ * ### Using loadFile
+ * ```ts
+ * const registry = new LoaderRegistry();
+ * const doc = await registry.loadFile('/notes/meeting.md');
+ * console.log(doc.metadata.title);
+ * ```
+ */
+export declare class LoaderRegistry {
+    /**
+     * Internal map from lower-cased extension (with dot) to the loader
+     * responsible for that extension.
+     *
+     * When multiple loaders claim the same extension the last one registered
+     * wins (newest-registration-wins semantics), allowing callers to override
+     * built-in loaders.
+     */
+    private readonly _loaders;
+    /**
+     * Creates a new registry pre-populated with the built-in loaders.
+     *
+     * Loader registration order determines conflict resolution: later
+     * registrations override earlier ones for the same extension.
+     *
+     * Registration order:
+     * 1. {@link TextLoader}, {@link MarkdownLoader}, {@link HtmlLoader} — core text formats.
+     * 2. {@link PdfLoader} (with injected OCR + Docling loaders) — PDF extraction.
+     * 3. {@link DocxLoader} — DOCX extraction.
+     * 4. Optional: an override from {@link createOcrPdfLoader} when `tesseract.js` is installed.
+     * 5. Optional: an override from {@link createDoclingLoader} when Python Docling is available.
+     *    The Docling-backed loader supports both `.pdf` and `.docx`, so it supersedes both
+     *    PdfLoader and DocxLoader when present.
+     */
+    constructor();
+    /**
+     * Register a loader for all extensions it declares.
+     *
+     * If a previously registered loader already handles one of the extension,
+     * it is replaced.  This makes it trivial to swap in a higher-fidelity
+     * implementation for any format.
+     *
+     * @param loader - The loader instance to register.
+     *
+     * @example
+     * ```ts
+     * registry.register(new PdfLoader());
+     * ```
+     */
+    register(loader: IDocumentLoader): void;
+    /**
+     * Retrieve the loader registered for `extensionOrPath`.
+     *
+     * Both bare extensions (`.md`, `md`) and full file paths
+     * (`/docs/guide.md`) are accepted.
+     *
+     * @param extensionOrPath - File extension or full path.
+     * @returns The matching {@link IDocumentLoader}, or `undefined` when no
+     *          loader is registered for the detected extension.
+     *
+     * @example
+     * ```ts
+     * const loader = registry.getLoader('.md');
+     * const loader2 = registry.getLoader('README.md');
+     * ```
+     */
+    getLoader(extensionOrPath: string): IDocumentLoader | undefined;
+    /**
+     * Return a sorted array of all extensions currently registered.
+     *
+     * Each extension is returned with a leading dot in lower-case, e.g.
+     * `['.csv', '.htm', '.html', '.json', '.md', …]`.
+     *
+     * @returns Sorted array of registered extension strings.
+     */
+    getSupportedExtensions(): string[];
+    /**
+     * Convenience method: detect format from `filePath`, find the matching
+     * loader, and delegate to its `load()` method.
+     *
+     * @param filePath - Absolute (or resolvable relative) file path.
+     * @param options  - Optional load hints forwarded to the loader.
+     * @returns A promise resolving to the {@link LoadedDocument}.
+     *
+     * @throws {Error} When no loader is registered for the file's extension.
+     * @throws {Error} When the underlying loader's `load()` throws.
+     *
+     * @example
+     * ```ts
+     * const doc = await registry.loadFile('/notes/architecture.md');
+     * ```
+     */
+    loadFile(filePath: string, options?: LoadOptions): Promise<LoadedDocument>;
+}
+//# sourceMappingURL=LoaderRegistry.d.ts.map

package/dist/cognition/memory/io/ingestion/LoaderRegistry.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"LoaderRegistry.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/LoaderRegistry.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAGH,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC5D,OAAO,KAAK,EAAE,WAAW,EAAE,cAAc,EAAE,MAAM,oBAAoB,CAAC;AAgDtE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAgCG;AACH,qBAAa,cAAc;IACzB;;;;;;;OAOG;IACH,OAAO,CAAC,QAAQ,CAAC,QAAQ,CAA2C;IAEpE;;;;;;;;;;;;;;OAcG;;IA6BH;;;;;;;;;;;;;OAaG;IACH,QAAQ,CAAC,MAAM,EAAE,eAAe,GAAG,IAAI;IAUvC;;;;;;;;;;;;;;;OAeG;IACH,SAAS,CAAC,eAAe,EAAE,MAAM,GAAG,eAAe,GAAG,SAAS;IAS/D;;;;;;;OAOG;IACH,sBAAsB,IAAI,MAAM,EAAE;IAQlC;;;;;;;;;;;;;;;OAeG;IACG,QAAQ,CAAC,QAAQ,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,cAAc,CAAC;CAajF"}

package/dist/cognition/memory/io/ingestion/LoaderRegistry.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"LoaderRegistry.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/LoaderRegistry.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAEH,OAAO,IAAI,MAAM,WAAW,CAAC;AAG7B,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,OAAO,EAAE,cAAc,EAAE,MAAM,qBAAqB,CAAC;AACrD,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,OAAO,EAAE,SAAS,EAAE,MAAM,gBAAgB,CAAC;AAC3C,OAAO,EAAE,UAAU,EAAE,MAAM,iBAAiB,CAAC;AAC7C,OAAO,EAAE,kBAAkB,EAAE,MAAM,mBAAmB,CAAC;AACvD,OAAO,EAAE,mBAAmB,EAAE,MAAM,oBAAoB,CAAC;AAEzD,8EAA8E;AAC9E,UAAU;AACV,8EAA8E;AAE9E;;;;;;;;;;;;;;GAcG;AACH,SAAS,YAAY,CAAC,eAAuB;IAC3C,4EAA4E;IAC5E,2EAA2E;IAC3E,0DAA0D;IAC1D,MAAM,YAAY,GAAG,eAAe,CAAC,QAAQ,CAAC,GAAG,CAAC,IAAI,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;IACrF,MAAM,cAAc,GAAG,eAAe,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;IAE5D,IAAI,CAAC,YAAY,IAAI,CAAC,cAAc,EAAE,CAAC;QACrC,uCAAuC;QACvC,MAAM,QAAQ,GAAG,eAAe,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,eAAe,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,eAAe,CAAC;QAC9F,OAAO,QAAQ,CAAC,CAAC,CAAC,IAAI,QAAQ,CAAC,WAAW,EAAE,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;IACtD,CAAC;IAED,OAAO,IAAI,CAAC,OAAO,CAAC,eAAe,CAAC,CAAC,WAAW,EAAE,CAAC;AACrD,CAAC;AAED,8EAA8E;AAC9E,iBAAiB;AACjB,8EAA8E;AAE9E;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAgCG;AACH,MAAM,OAAO,cAAc;IAWzB;;;;;;;;;;;;;;OAcG;IACH;QAzBA;;;;;;;WAOG;QACc,aAAQ,GAAiC,IAAI,GAAG,EAAE,CAAC;QAkBlE,4BAA4B;QAC5B,IAAI,CAAC,QAAQ,CAAC,IAAI,UAAU,EAAE,CAAC,CAAC;QAChC,IAAI,CAAC,QAAQ,CAAC,IAAI,cAAc,EAAE,CAAC,CAAC;QACpC,IAAI,CAAC,QAAQ,CAAC,IAAI,UAAU,EAAE,CAAC,CAAC;QAEhC,wEAAwE;QACxE,0EAA0E;QAC1E,MAAM,SAAS,GAAG,kBAAkB,EAAE,CAAC;QACvC,MAAM,aAAa,GAAG,mBAAmB,EAAE,CAAC;QAE5C,+DAA+D;QAC/D,IAAI,CAAC,QAAQ,CAAC,IAAI,SAAS,CAAC,SAAS,EAAE,aAAa,CAAC,CAAC,CAAC;QAEvD,eAAe;QACf,IAAI,CAAC,QAAQ,CAAC,IAAI,UAAU,EAAE,CAAC,CAAC;QAEhC,wEAAwE;QACxE,6DAA6D;QAC7D,IAAI,aAAa,KAAK,IAAI,EAAE,CAAC;YAC3B,IAAI,CAAC,QAAQ,CAAC,aAAa,CAAC,CAAC;QAC/B,CAAC;IACH,CAAC;IAED,4EAA4E;IAC5E,WAAW;IACX,4EAA4E;IAE5E;;;;;;;;;;;;;OAaG;IACH,QAAQ,CAAC,MAAuB;QAC9B,KAAK,MAAM,GAAG,IAAI,MAAM,CAAC,mBAAmB,EAAE,CAAC;YAC7C,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAC,WAAW,EAAE,EAAE,MAAM,CAAC,CAAC;QAC/C,CAAC;IACH,CAAC;IAED,4EAA4E;IAC5E,YAAY;IACZ,4EAA4E;IAE5E;;;;;;;;;;;;;;;OAeG;IACH,SAAS,CAAC,eAAuB;QAC/B,MAAM,GAAG,GAAG,YAAY,CAAC,eAAe,CAAC,CAAC;QAC1C,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;IAChC,CAAC;IAED,4EAA4E;IAC5E,yBAAyB;IACzB,4EAA4E;IAE5E;;;;;;;OAOG;IACH,sBAAsB;QACpB,OAAO,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;IAC1C,CAAC;IAED,4EAA4E;IAC5E,WAAW;IACX,4EAA4E;IAE5E;;;;;;;;;;;;;;;OAeG;IACH,KAAK,CAAC,QAAQ,CAAC,QAAgB,EAAE,OAAqB;QACpD,MAAM,MAAM,GAAG,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC;QAExC,IAAI,CAAC,MAAM,EAAE,CAAC;YACZ,MAAM,GAAG,GAAG,YAAY,CAAC,QAAQ,CAAC,CAAC;YACnC,MAAM,IAAI,KAAK,CACb,uDAAuD,GAAG,aAAa,QAAQ,MAAM;gBACnF,yBAAyB,IAAI,CAAC,sBAAsB,EAAE,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,CACvE,CAAC;QACJ,CAAC;QAED,OAAO,MAAM,CAAC,IAAI,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;IACxC,CAAC;CACF"}

package/dist/cognition/memory/io/ingestion/MarkdownLoader.d.ts

@@ -0,0 +1,50 @@
+/**
+ * @fileoverview MarkdownLoader — loads `.md` and `.mdx` documents.
+ *
+ * Parses YAML front-matter using the `gray-matter` library, strips it from
+ * the returned content, and promotes key metadata fields (title, author,
+ * createdAt, etc.) into the {@link DocumentMetadata} shape.
+ *
+ * When no `title` key is present in the front-matter the loader falls back
+ * to extracting the first ATX heading (`# …`) from the document body.
+ *
+ * @module memory/ingestion/MarkdownLoader
+ */
+import type { IDocumentLoader } from './IDocumentLoader.js';
+import type { LoadOptions, LoadedDocument } from '../facade/types.js';
+/**
+ * Document loader for Markdown (`.md`) and MDX (`.mdx`) files.
+ *
+ * ### Front-matter handling
+ * YAML front-matter delimited by `---` is parsed via `gray-matter`.  All
+ * key-value pairs are merged into {@link DocumentMetadata} as-is, with a
+ * handful of well-known keys (`title`, `author`, `createdAt`, `modifiedAt`,
+ * `language`) mapped to the corresponding typed metadata fields.
+ *
+ * ### Title extraction fallback
+ * When the front-matter does **not** contain a `title` field the loader
+ * searches the document body for the first level-1 ATX heading (`# Title`)
+ * and uses that as the title.
+ *
+ * ### Returned content
+ * The `content` field in the returned {@link LoadedDocument} contains the
+ * Markdown body **without** the front-matter block.
+ *
+ * @implements {IDocumentLoader}
+ *
+ * @example
+ * ```ts
+ * const loader = new MarkdownLoader();
+ * const doc = await loader.load('/docs/architecture.md');
+ * console.log(doc.metadata.title); // from front-matter or first # heading
+ * ```
+ */
+export declare class MarkdownLoader implements IDocumentLoader {
+    /** @inheritdoc */
+    readonly supportedExtensions: string[];
+    /** @inheritdoc */
+    canLoad(source: string | Buffer): boolean;
+    /** @inheritdoc */
+    load(source: string | Buffer, _options?: LoadOptions): Promise<LoadedDocument>;
+}
+//# sourceMappingURL=MarkdownLoader.d.ts.map

package/dist/cognition/memory/io/ingestion/MarkdownLoader.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"MarkdownLoader.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/MarkdownLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;GAWG;AAKH,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC5D,OAAO,KAAK,EAAE,WAAW,EAAE,cAAc,EAAoB,MAAM,oBAAoB,CAAC;AAqExF;;;;;;;;;;;;;;;;;;;;;;;;;;GA0BG;AACH,qBAAa,cAAe,YAAW,eAAe;IACpD,kBAAkB;IAClB,QAAQ,CAAC,mBAAmB,EAAE,MAAM,EAAE,CAA6B;IAMnE,kBAAkB;IAClB,OAAO,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,GAAG,OAAO;IAazC,kBAAkB;IACZ,IAAI,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,QAAQ,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,cAAc,CAAC;CAqDrF"}

package/dist/cognition/memory/io/ingestion/MarkdownLoader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"MarkdownLoader.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/MarkdownLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;GAWG;AAEH,OAAO,EAAE,MAAM,kBAAkB,CAAC;AAClC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,MAAM,MAAM,aAAa,CAAC;AAGjC,OAAO,EAAE,YAAY,EAAE,MAAM,gBAAgB,CAAC;AAE9C,8EAA8E;AAC9E,YAAY;AACZ,8EAA8E;AAE9E,kEAAkE;AAClE,MAAM,oBAAoB,GAAG,CAAC,KAAK,EAAE,MAAM,CAAU,CAAC;AAEtD,8EAA8E;AAC9E,UAAU;AACV,8EAA8E;AAE9E;;;;GAIG;AACH,SAAS,KAAK,CAAC,QAAgB;IAC7B,OAAO,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;AAC9C,CAAC;AAED;;;;GAIG;AACH,SAAS,SAAS,CAAC,IAAY;IAC7B,OAAO,IAAI,CAAC,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC;AAClE,CAAC;AAED;;;;;;;;GAQG;AACH,SAAS,mBAAmB,CAAC,IAAY;IACvC,8EAA8E;IAC9E,MAAM,KAAK,GAAG,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACzC,OAAO,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC;AAC7C,CAAC;AAED;;;;;;GAMG;AACH,SAAS,WAAW,CAAC,KAAc;IACjC,IAAI,KAAK,YAAY,IAAI,EAAE,CAAC;QAC1B,OAAO,KAAK,CAAC,WAAW,EAAE,CAAC;IAC7B,CAAC;IACD,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,OAAO,KAAK,KAAK,QAAQ,EAAE,CAAC;QAC3D,MAAM,CAAC,GAAG,IAAI,IAAI,CAAC,KAAK,CAAC,CAAC;QAC1B,OAAO,KAAK,CAAC,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC;IAC1D,CAAC;IACD,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,8EAA8E;AAC9E,iBAAiB;AACjB,8EAA8E;AAE9E;;;;;;;;;;;;;;;;;;;;;;;;;;GA0BG;AACH,MAAM,OAAO,cAAc;IAA3B;QACE,kBAAkB;QACT,wBAAmB,GAAa,CAAC,GAAG,oBAAoB,CAAC,CAAC;IA0ErE,CAAC;IAxEC,4EAA4E;IAC5E,UAAU;IACV,4EAA4E;IAE5E,kBAAkB;IAClB,OAAO,CAAC,MAAuB;QAC7B,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,sEAAsE;YACtE,sEAAsE;YACtE,OAAO,KAAK,CAAC;QACf,CAAC;QACD,OAAQ,oBAA0C,CAAC,QAAQ,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;IAC7E,CAAC;IAED,4EAA4E;IAC5E,OAAO;IACP,4EAA4E;IAE5E,kBAAkB;IAClB,KAAK,CAAC,IAAI,CAAC,MAAuB,EAAE,QAAsB;QACxD,IAAI,GAAW,CAAC;QAChB,IAAI,YAAgC,CAAC;QAErC,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,GAAG,GAAG,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;QAChC,CAAC;aAAM,CAAC;YACN,YAAY,GAAG,YAAY,CAAC,MAAM,CAAC,CAAC;YACpC,MAAM,KAAK,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC;YAC9C,GAAG,GAAG,KAAK,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;QAC/B,CAAC;QAED,+BAA+B;QAC/B,MAAM,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC;QAE3B,2DAA2D;QAC3D,MAAM,IAAI,GAAG,MAAM,CAAC,OAAO,CAAC;QAE5B,2BAA2B;QAC3B,MAAM,EAAE,GAAG,MAAM,CAAC,IAA+B,CAAC;QAElD,gEAAgE;QAChE,MAAM,OAAO,GACX,OAAO,EAAE,CAAC,OAAO,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QAC5D,MAAM,YAAY,GAAG,OAAO,KAAK,SAAS,CAAC,CAAC,CAAC,mBAAmB,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QACnF,MAAM,KAAK,GAAG,OAAO,IAAI,YAAY,CAAC;QAEtC,MAAM,IAAI,GAAqB;YAC7B,4BAA4B;YAC5B,GAAG,CAAC,KAAK,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YACzC,GAAG,CAAC,OAAO,EAAE,CAAC,QAAQ,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,EAAE,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YACrE,GAAG,CAAC,EAAE,CAAC,WAAW,CAAC,KAAK,SAAS;gBAC/B,CAAC,CAAC,EAAE,SAAS,EAAE,WAAW,CAAC,EAAE,CAAC,WAAW,CAAC,CAAC,EAAE;gBAC7C,CAAC,CAAC,EAAE,CAAC;YACP,GAAG,CAAC,EAAE,CAAC,YAAY,CAAC,KAAK,SAAS;gBAChC,CAAC,CAAC,EAAE,UAAU,EAAE,WAAW,CAAC,EAAE,CAAC,YAAY,CAAC,CAAC,EAAE;gBAC/C,CAAC,CAAC,EAAE,CAAC;YACP,GAAG,CAAC,OAAO,EAAE,CAAC,UAAU,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,EAAE,CAAC,UAAU,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC3E,4DAA4D;YAC5D,GAAG,EAAE;YACL,oEAAoE;YACpE,mDAAmD;YACnD,GAAG,CAAC,KAAK,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YACzC,SAAS,EAAE,SAAS,CAAC,IAAI,CAAC;YAC1B,GAAG,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SAClD,CAAC;QAEF,OAAO;YACL,OAAO,EAAE,IAAI;YACb,QAAQ,EAAE,IAAI;YACd,MAAM,EAAE,IAAI;SACb,CAAC;IACJ,CAAC;CACF"}

package/dist/cognition/memory/io/ingestion/MultimodalAggregator.d.ts

@@ -0,0 +1,88 @@
+/**
+ * @fileoverview MultimodalAggregator — post-processing stage for images
+ * extracted from documents.
+ *
+ * After loaders such as {@link PdfLoader} and {@link DocxLoader} extract
+ * embedded images as {@link ExtractedImage} objects, `MultimodalAggregator`
+ * enriches them with natural-language captions by optionally calling a
+ * vision-capable LLM function supplied by the application layer.
+ *
+ * The class is intentionally thin: it holds no state beyond the optional
+ * configuration and delegates all vision intelligence to the caller-supplied
+ * `describeImage` function.  This keeps the aggregator testable without any
+ * live LLM dependencies.
+ *
+ * @module memory/ingestion/MultimodalAggregator
+ */
+import type { ExtractedImage } from '../facade/types.js';
+/**
+ * Configuration for {@link MultimodalAggregator}.
+ */
+export interface MultimodalConfig {
+    /**
+     * Async function that accepts a raw image buffer and its MIME type and
+     * returns a natural-language description of the image.
+     *
+     * When this is `undefined` the aggregator passes images through unchanged.
+     *
+     * @param imageBuffer - Raw bytes of the image (PNG, JPEG, WebP, …).
+     * @param mimeType    - MIME type of the image, e.g. `'image/png'`.
+     * @returns A promise resolving to a human-readable description string.
+     *
+     * @example
+     * ```ts
+     * async (buffer, mimeType) => {
+     *   return openaiClient.vision(buffer, mimeType);
+     * }
+     * ```
+     */
+    describeImage?: (imageBuffer: Buffer, mimeType: string) => Promise<string>;
+}
+/**
+ * Adds auto-generated captions to {@link ExtractedImage} objects that lack
+ * one, using a caller-supplied vision LLM function.
+ *
+ * Images are processed in parallel via `Promise.allSettled()` so a single
+ * failed captioning attempt does not block the rest.  Images whose captioning
+ * fails retain their original (un-captioned) state rather than propagating the
+ * error.
+ *
+ * ### Example — with a vision LLM
+ * ```ts
+ * const aggregator = new MultimodalAggregator({
+ *   describeImage: async (buf, mime) => myVisionLLM.describe(buf, mime),
+ * });
+ *
+ * const captioned = await aggregator.processImages(doc.images ?? []);
+ * ```
+ *
+ * ### Example — passthrough (no LLM configured)
+ * ```ts
+ * const aggregator = new MultimodalAggregator();
+ * const unchanged  = await aggregator.processImages(doc.images ?? []);
+ * ```
+ */
+export declare class MultimodalAggregator {
+    private readonly config?;
+    /**
+     * @param config - Optional configuration.  Omit to use in passthrough mode.
+     */
+    constructor(config?: MultimodalConfig | undefined);
+    /**
+     * Enrich images with captions via the configured vision LLM.
+     *
+     * Only images that have no existing `caption` field are processed.  Images
+     * that already carry a caption are left unchanged to avoid redundant LLM
+     * calls.
+     *
+     * When no `describeImage` function is configured all images are returned
+     * unchanged.
+     *
+     * @param images - Array of {@link ExtractedImage} objects to process.
+     * @returns A promise resolving to the same-length array of
+     *          {@link ExtractedImage} objects, with captions filled in where
+     *          possible.
+     */
+    processImages(images: ExtractedImage[]): Promise<ExtractedImage[]>;
+}
+//# sourceMappingURL=MultimodalAggregator.d.ts.map

package/dist/cognition/memory/io/ingestion/MultimodalAggregator.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"MultimodalAggregator.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/MultimodalAggregator.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAEH,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,oBAAoB,CAAC;AAMzD;;GAEG;AACH,MAAM,WAAW,gBAAgB;IAC/B;;;;;;;;;;;;;;;;OAgBG;IACH,aAAa,CAAC,EAAE,CAAC,WAAW,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;CAC5E;AAMD;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AACH,qBAAa,oBAAoB;IAInB,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAC;IAHpC;;OAEG;gBAC0B,MAAM,CAAC,EAAE,gBAAgB,YAAA;IAMtD;;;;;;;;;;;;;;OAcG;IACG,aAAa,CAAC,MAAM,EAAE,cAAc,EAAE,GAAG,OAAO,CAAC,cAAc,EAAE,CAAC;CAiCzE"}

package/dist/cognition/memory/io/ingestion/MultimodalAggregator.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"MultimodalAggregator.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/MultimodalAggregator.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAgCH,8EAA8E;AAC9E,uBAAuB;AACvB,8EAA8E;AAE9E;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AACH,MAAM,OAAO,oBAAoB;IAC/B;;OAEG;IACH,YAA6B,MAAyB;QAAzB,WAAM,GAAN,MAAM,CAAmB;IAAG,CAAC;IAE1D,4EAA4E;IAC5E,gBAAgB;IAChB,4EAA4E;IAE5E;;;;;;;;;;;;;;OAcG;IACH,KAAK,CAAC,aAAa,CAAC,MAAwB;QAC1C,0EAA0E;QAC1E,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,aAAa,EAAE,CAAC;YAChC,OAAO,MAAM,CAAC,KAAK,EAAE,CAAC;QACxB,CAAC;QAED,MAAM,aAAa,GAAG,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC;QAEhD,gEAAgE;QAChE,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,UAAU,CACtC,MAAM,CAAC,GAAG,CAAC,KAAK,EAAE,KAAK,EAA2B,EAAE;YAClD,2CAA2C;YAC3C,IAAI,KAAK,CAAC,OAAO,KAAK,SAAS,EAAE,CAAC;gBAChC,OAAO,KAAK,CAAC;YACf,CAAC;YAED,IAAI,CAAC;gBACH,MAAM,OAAO,GAAG,MAAM,aAAa,CAAC,KAAK,CAAC,IAAI,EAAE,KAAK,CAAC,QAAQ,CAAC,CAAC;gBAChE,OAAO,EAAE,GAAG,KAAK,EAAE,OAAO,EAAE,CAAC;YAC/B,CAAC;YAAC,MAAM,CAAC;gBACP,2DAA2D;gBAC3D,OAAO,KAAK,CAAC;YACf,CAAC;QACH,CAAC,CAAC,CACH,CAAC;QAEF,yEAAyE;QACzE,2EAA2E;QAC3E,uEAAuE;QACvE,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,KAAK,EAAE,EAAE,CACnC,MAAM,CAAC,MAAM,KAAK,WAAW,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAC7D,CAAC;IACJ,CAAC;CACF"}

package/dist/cognition/memory/io/ingestion/OcrPdfLoader.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"OcrPdfLoader.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/OcrPdfLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;GAiBG;AAKH,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAsH5D;;;;;;;;;;;;;;;;;;;GAmBG;AACH,wBAAgB,kBAAkB,IAAI,eAAe,GAAG,IAAI,CAY3D"}

package/dist/cognition/memory/io/ingestion/OcrPdfLoader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"OcrPdfLoader.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/OcrPdfLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;GAiBG;AAEH,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,MAAM,kBAAkB,CAAC;AAClC,OAAO,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAI5C,8EAA8E;AAC9E,YAAY;AACZ,8EAA8E;AAE9E,sCAAsC;AACtC,MAAM,oBAAoB,GAAG,CAAC,MAAM,CAAU,CAAC;AAE/C,8EAA8E;AAC9E,UAAU;AACV,8EAA8E;AAE9E;;;;GAIG;AACH,SAAS,KAAK,CAAC,QAAgB;IAC7B,OAAO,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;AAC9C,CAAC;AAED,8EAA8E;AAC9E,gCAAgC;AAChC,8EAA8E;AAE9E;;;;;;GAMG;AACH,MAAM,YAAY;IAAlB;QACE,kBAAkB;QACT,wBAAmB,GAAa,CAAC,GAAG,oBAAoB,CAAC,CAAC;IA4ErE,CAAC;IA1EC,4EAA4E;IAC5E,UAAU;IACV,4EAA4E;IAE5E,kBAAkB;IAClB,OAAO,CAAC,MAAuB;QAC7B,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,6CAA6C;YAC7C,OAAO,CACL,MAAM,CAAC,MAAM,IAAI,CAAC;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,CACnB,CAAC;QACJ,CAAC;QACD,OAAQ,oBAA0C,CAAC,QAAQ,CAAC,KAAK,CAAC,MAAM,CAAW,CAAC,CAAC;IACvF,CAAC;IAED,4EAA4E;IAC5E,OAAO;IACP,4EAA4E;IAE5E,kBAAkB;IAClB,KAAK,CAAC,IAAI,CAAC,MAAuB,EAAE,QAAsB;QACxD,IAAI,MAAc,CAAC;QACnB,IAAI,YAAgC,CAAC;QAErC,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,MAAM,GAAG,MAAM,CAAC;QAClB,CAAC;aAAM,CAAC;YACN,YAAY,GAAG,MAAM,CAAC;YACtB,MAAM,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC;QAC3C,CAAC;QAED,wEAAwE;QACxE,sEAAsE;QACtE,uEAAuE;QACvE,yEAAyE;QACzE,8DAA8D;QAC9D,IAAI,SAAc,CAAC;QACnB,IAAI,CAAC;YACH,6DAA6D;YAC7D,8EAA8E;YAC9E,SAAS,GAAG,MAAM,MAAM,CAAC,cAAc,CAAC,CAAC;QAC3C,CAAC;QAAC,MAAM,CAAC;YACP,MAAM,IAAI,KAAK,CACb,gDAAgD;gBAC9C,2EAA2E;gBAC3E,qCAAqC,CACxC,CAAC;QACJ,CAAC;QAED,oCAAoC;QACpC,oEAAoE;QACpE,2EAA2E;QAC3E,MAAM,EACJ,IAAI,EAAE,EAAE,IAAI,EAAE,GACf,GAAG,MAAM,SAAS,CAAC,SAAS,CAAC,MAAM,EAAE,KAAK,EAAE;YAC3C,MAAM,EAAE,GAAG,EAAE;gBACX,8BAA8B;YAChC,CAAC;SACF,CAAC,CAAC;QAEH,MAAM,IAAI,GAAqB;YAC7B,GAAG,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SAClD,CAAC;QAEF,OAAO;YACL,OAAO,EAAE,IAAI;YACb,QAAQ,EAAE,IAAI;YACd,MAAM,EAAE,KAAK;SACd,CAAC;IACJ,CAAC;CACF;AAED,8EAA8E;AAC9E,UAAU;AACV,8EAA8E;AAE9E;;;;;;;;;;;;;;;;;;;GAmBG;AACH,MAAM,UAAU,kBAAkB;IAChC,IAAI,CAAC;QACH,0EAA0E;QAC1E,4EAA4E;QAC5E,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC/C,OAAO,CAAC,OAAO,CAAC,cAAc,CAAC,CAAC;QAChC,oDAAoD;QACpD,OAAO,IAAI,YAAY,EAAE,CAAC;IAC5B,CAAC;IAAC,MAAM,CAAC;QACP,8CAA8C;QAC9C,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC"}

package/dist/cognition/memory/io/ingestion/PdfLoader.d.ts

@@ -0,0 +1,78 @@
+/**
+ * @fileoverview PdfLoader — loads `.pdf` documents using `unpdf`.
+ *
+ * Implements a tiered extraction strategy:
+ * 1. **unpdf** (Tier 1, always available) — pure-JS text extraction via
+ *    `getDocumentProxy` + `extractText`.  Fast and dependency-free.
+ * 2. **OCR fallback** (Tier 2, opt-in) — Tesseract.js OCR, engaged when the
+ *    unpdf extraction produces sparse text (< 50 chars per page on average).
+ * 3. **Docling fallback** (Tier 3, opt-in) — Python `docling` subprocess, engaged
+ *    when explicitly configured via the `docling` loader option.
+ *
+ * Both fallback loaders are optional and injected via constructor parameters;
+ * callers supply them by calling `createOcrPdfLoader()` and
+ * `createDoclingLoader()` and checking for non-null values.
+ *
+ * @module memory/ingestion/PdfLoader
+ */
+import type { IDocumentLoader } from './IDocumentLoader.js';
+import type { LoadOptions, LoadedDocument } from '../facade/types.js';
+/**
+ * Document loader for PDF files.
+ *
+ * ### Extraction tiers
+ * 1. **unpdf** — always used as the primary extraction engine.  Performs
+ *    pure-JS PDF text layer extraction with no native binaries required.
+ * 2. **OCR fallback** (optional) — supplied at construction time and engaged
+ *    automatically when unpdf yields sparse text (< 50 chars per page on
+ *    average), indicating a scanned document.
+ * 3. **Docling fallback** (optional) — when provided, takes precedence over both
+ *    unpdf and OCR, yielding the highest-fidelity extraction at the cost of
+ *    requiring a Python runtime.
+ *
+ * @implements {IDocumentLoader}
+ *
+ * @example
+ * ```ts
+ * const ocrLoader    = createOcrPdfLoader();   // null if tesseract.js absent
+ * const doclingLoader = createDoclingLoader(); // null if docling absent
+ * const pdfLoader = new PdfLoader(ocrLoader, doclingLoader);
+ * const doc = await pdfLoader.load('/reports/q3.pdf');
+ * ```
+ */
+export declare class PdfLoader implements IDocumentLoader {
+    /** @inheritdoc */
+    readonly supportedExtensions: string[];
+    /**
+     * Optional OCR fallback loader, engaged when primary extraction is sparse.
+     * Pass `null` to disable OCR fallback.
+     */
+    private readonly _ocrLoader;
+    /**
+     * Optional Docling loader that, when present, takes precedence over the
+     * entire unpdf + OCR pipeline.
+     * Pass `null` to disable Docling.
+     */
+    private readonly _doclingLoader;
+    /**
+     * Creates a new PdfLoader.
+     *
+     * @param ocrLoader - Optional OCR fallback (for example from `createOcrPdfLoader()`).
+     * @param doclingLoader - Optional Docling loader (for example from `createDoclingLoader()`).
+     */
+    constructor(ocrLoader?: IDocumentLoader | null, doclingLoader?: IDocumentLoader | null);
+    /** @inheritdoc */
+    canLoad(source: string | Buffer): boolean;
+    /** @inheritdoc */
+    load(source: string | Buffer, options?: LoadOptions): Promise<LoadedDocument>;
+    /**
+     * Extract text from a PDF buffer using the `unpdf` package.
+     *
+     * Returns the concatenated page text, the page count, and an optional title
+     * string extracted from the PDF metadata dictionary when available.
+     *
+     * @param buffer - Raw PDF bytes.
+     */
+    private _extractWithUnpdf;
+}
+//# sourceMappingURL=PdfLoader.d.ts.map

package/dist/cognition/memory/io/ingestion/PdfLoader.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"PdfLoader.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/PdfLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAIH,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC5D,OAAO,KAAK,EAAE,WAAW,EAAE,cAAc,EAAoB,MAAM,oBAAoB,CAAC;AAgDxF;;;;;;;;;;;;;;;;;;;;;;GAsBG;AACH,qBAAa,SAAU,YAAW,eAAe;IAC/C,kBAAkB;IAClB,QAAQ,CAAC,mBAAmB,EAAE,MAAM,EAAE,CAA6B;IAEnE;;;OAGG;IACH,OAAO,CAAC,QAAQ,CAAC,UAAU,CAAyB;IAEpD;;;;OAIG;IACH,OAAO,CAAC,QAAQ,CAAC,cAAc,CAAyB;IAExD;;;;;OAKG;gBAED,SAAS,GAAE,eAAe,GAAG,IAAW,EACxC,aAAa,GAAE,eAAe,GAAG,IAAW;IAU9C,kBAAkB;IAClB,OAAO,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,GAAG,OAAO;IAkBzC,kBAAkB;IACZ,IAAI,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,OAAO,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,cAAc,CAAC;IA2CnF;;;;;;;OAOG;YACW,iBAAiB;CA+BhC"}

package/dist/cognition/memory/io/ingestion/PdfLoader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"PdfLoader.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/PdfLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,EAAE,MAAM,kBAAkB,CAAC;AAClC,OAAO,IAAI,MAAM,WAAW,CAAC;AAG7B,OAAO,EAAE,YAAY,EAAE,MAAM,gBAAgB,CAAC;AAE9C,8EAA8E;AAC9E,YAAY;AACZ,8EAA8E;AAE9E,kEAAkE;AAClE,MAAM,oBAAoB,GAAG,CAAC,MAAM,CAAU,CAAC;AAE/C;;;GAGG;AACH,MAAM,+BAA+B,GAAG,EAAE,CAAC;AAE3C,8EAA8E;AAC9E,UAAU;AACV,8EAA8E;AAE9E;;;;GAIG;AACH,SAAS,KAAK,CAAC,QAAgB;IAC7B,OAAO,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;AAC9C,CAAC;AAED;;;;;;;;GAQG;AACH,SAAS,QAAQ,CAAC,IAAY,EAAE,SAAiB;IAC/C,IAAI,SAAS,IAAI,CAAC;QAAE,OAAO,KAAK,CAAC;IACjC,MAAM,eAAe,GAAG,IAAI,CAAC,MAAM,GAAG,SAAS,CAAC;IAChD,OAAO,eAAe,GAAG,+BAA+B,CAAC;AAC3D,CAAC;AAED,8EAA8E;AAC9E,YAAY;AACZ,8EAA8E;AAE9E;;;;;;;;;;;;;;;;;;;;;;GAsBG;AACH,MAAM,OAAO,SAAS;IAiBpB;;;;;OAKG;IACH,YACE,YAAoC,IAAI,EACxC,gBAAwC,IAAI;QAxB9C,kBAAkB;QACT,wBAAmB,GAAa,CAAC,GAAG,oBAAoB,CAAC,CAAC;QAyBjE,IAAI,CAAC,UAAU,GAAG,SAAS,CAAC;QAC5B,IAAI,CAAC,cAAc,GAAG,aAAa,CAAC;IACtC,CAAC;IAED,4EAA4E;IAC5E,UAAU;IACV,4EAA4E;IAE5E,kBAAkB;IAClB,OAAO,CAAC,MAAuB;QAC7B,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,6CAA6C;YAC7C,OAAO,CACL,MAAM,CAAC,MAAM,IAAI,CAAC;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI;gBAClB,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,CACnB,CAAC;QACJ,CAAC;QACD,OAAQ,oBAA0C,CAAC,QAAQ,CAAC,KAAK,CAAC,MAAM,CAAW,CAAC,CAAC;IACvF,CAAC;IAED,4EAA4E;IAC5E,OAAO;IACP,4EAA4E;IAE5E,kBAAkB;IAClB,KAAK,CAAC,IAAI,CAAC,MAAuB,EAAE,OAAqB;QACvD,oDAAoD;QACpD,IAAI,IAAI,CAAC,cAAc,KAAK,IAAI,EAAE,CAAC;YACjC,OAAO,IAAI,CAAC,cAAc,CAAC,IAAI,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;QACnD,CAAC;QAED,+CAA+C;QAC/C,IAAI,MAAc,CAAC;QACnB,IAAI,YAAgC,CAAC;QAErC,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,MAAM,GAAG,MAAM,CAAC;QAClB,CAAC;aAAM,CAAC;YACN,YAAY,GAAG,YAAY,CAAC,MAAM,CAAC,CAAC;YACpC,MAAM,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC;QAC3C,CAAC;QAED,gCAAgC;QAChC,MAAM,EAAE,OAAO,EAAE,SAAS,EAAE,KAAK,EAAE,GAAG,MAAM,IAAI,CAAC,iBAAiB,CAAC,MAAM,CAAC,CAAC;QAE3E,0EAA0E;QAC1E,6DAA6D;QAC7D,IAAI,QAAQ,CAAC,OAAO,EAAE,SAAS,CAAC,IAAI,IAAI,CAAC,UAAU,KAAK,IAAI,EAAE,CAAC;YAC7D,OAAO,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;QAC/C,CAAC;QAED,MAAM,IAAI,GAAqB;YAC7B,SAAS;YACT,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC3B,GAAG,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SAClD,CAAC;QAEF,OAAO;YACL,OAAO;YACP,QAAQ,EAAE,IAAI;YACd,MAAM,EAAE,KAAK;SACd,CAAC;IACJ,CAAC;IAED,4EAA4E;IAC5E,4BAA4B;IAC5B,4EAA4E;IAE5E;;;;;;;OAOG;IACK,KAAK,CAAC,iBAAiB,CAAC,MAAc;QAK5C,yEAAyE;QACzE,wDAAwD;QACxD,MAAM,EAAE,gBAAgB,EAAE,WAAW,EAAE,GAAG,MAAM,MAAM,CAAC,OAAO,CAAC,CAAC;QAEhE,gEAAgE;QAChE,MAAM,GAAG,GAAG,MAAM,gBAAgB,CAAC,IAAI,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC;QAE3D,6BAA6B;QAC7B,MAAM,EAAE,IAAI,EAAE,GAAG,MAAM,WAAW,CAAC,GAAG,EAAE,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC,CAAC;QAE9D,gEAAgE;QAChE,IAAI,KAAyB,CAAC;QAC9B,IAAI,CAAC;YACH,MAAM,QAAQ,GAAG,MAAM,GAAG,CAAC,WAAW,EAAE,CAAC;YACzC,MAAM,IAAI,GAAG,QAAQ,EAAE,IAA2C,CAAC;YACnE,IAAI,IAAI,IAAI,OAAO,IAAI,CAAC,OAAO,CAAC,KAAK,QAAQ,IAAI,IAAI,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC;gBACtE,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,CAAC;YAC/B,CAAC;QACH,CAAC;QAAC,MAAM,CAAC;YACP,yDAAyD;QAC3D,CAAC;QAED,MAAM,SAAS,GAAG,GAAG,CAAC,QAAQ,IAAI,CAAC,CAAC;QAEpC,OAAO,EAAE,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC;IAC7C,CAAC;CACF"}

package/dist/cognition/memory/io/ingestion/TextLoader.d.ts

@@ -0,0 +1,66 @@
+/**
+ * @fileoverview TextLoader — loads plain-text, CSV/TSV, JSON, and YAML files.
+ *
+ * This is the most general-purpose loader in the AgentOS ingestion pipeline.
+ * It handles six extensions that all share the same fundamental operation:
+ * read raw text and attach lightweight metadata derived from the file content
+ * and extension.
+ *
+ * Supported extensions: `.txt`, `.csv`, `.tsv`, `.json`, `.yaml`, `.yml`
+ *
+ * @module memory/ingestion/TextLoader
+ */
+import type { IDocumentLoader } from './IDocumentLoader.js';
+import type { LoadOptions, LoadedDocument } from '../facade/types.js';
+/**
+ * Loader for plain-text, CSV, TSV, JSON, and YAML files.
+ *
+ * The loader performs minimal transformation:
+ * - **`.json`** — re-serialises with pretty-printing so stored content is
+ *   consistently formatted.
+ * - **`.yaml` / `.yml`** — the `yaml` package is used to parse and re-dump
+ *   for consistent formatting; falls back to raw text on parse error.
+ * - All other extensions — content is returned as-is.
+ *
+ * Metadata includes the approximate `wordCount` and a `format` label derived
+ * from the file extension.
+ *
+ * @implements {IDocumentLoader}
+ *
+ * @example
+ * ```ts
+ * const loader = new TextLoader();
+ * const doc = await loader.load('/data/notes.txt');
+ * console.log(doc.metadata.wordCount); // e.g. 312
+ * ```
+ */
+export declare class TextLoader implements IDocumentLoader {
+    /** @inheritdoc */
+    readonly supportedExtensions: string[];
+    /** @inheritdoc */
+    canLoad(source: string | Buffer): boolean;
+    /** @inheritdoc */
+    load(source: string | Buffer, _options?: LoadOptions): Promise<LoadedDocument>;
+    /**
+     * Normalises raw file content based on the detected extension.
+     *
+     * - JSON files are pretty-printed.
+     * - YAML files are parsed and re-dumped for consistent formatting.
+     * - All other formats are returned unchanged.
+     *
+     * @param raw - Raw UTF-8 string read from the source.
+     * @param ext - Lower-cased extension with leading dot.
+     */
+    private _normalise;
+    /**
+     * Parse and re-serialise YAML content for consistent formatting.
+     *
+     * Uses the `yaml` package that is already a production dependency of the
+     * `@framers/agentos` package.  Falls back to the original raw string on
+     * any parse error so the loader never throws on malformed YAML.
+     *
+     * @param raw - Raw YAML string.
+     */
+    private _prettyYaml;
+}
+//# sourceMappingURL=TextLoader.d.ts.map

package/dist/cognition/memory/io/ingestion/TextLoader.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"TextLoader.d.ts","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/TextLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;GAWG;AAKH,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAC5D,OAAO,KAAK,EAAE,WAAW,EAAE,cAAc,EAAoB,MAAM,oBAAoB,CAAC;AAuFxF;;;;;;;;;;;;;;;;;;;;;GAqBG;AACH,qBAAa,UAAW,YAAW,eAAe;IAChD,kBAAkB;IAClB,QAAQ,CAAC,mBAAmB,EAAE,MAAM,EAAE,CAA6B;IAMnE,kBAAkB;IAClB,OAAO,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,GAAG,OAAO;IAazC,kBAAkB;IACZ,IAAI,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,QAAQ,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,cAAc,CAAC;IAsCpF;;;;;;;;;OASG;IACH,OAAO,CAAC,UAAU;IAelB;;;;;;;;OAQG;IACH,OAAO,CAAC,WAAW;CAQpB"}

package/dist/cognition/memory/io/ingestion/TextLoader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"TextLoader.js","sourceRoot":"","sources":["../../../../../src/cognition/memory/io/ingestion/TextLoader.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;GAWG;AAEH,OAAO,EAAE,MAAM,kBAAkB,CAAC;AAClC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,KAAK,IAAI,SAAS,EAAE,SAAS,IAAI,aAAa,EAAE,MAAM,MAAM,CAAC;AAGtE,OAAO,EAAE,YAAY,EAAE,MAAM,gBAAgB,CAAC;AAE9C,8EAA8E;AAC9E,YAAY;AACZ,8EAA8E;AAE9E,6EAA6E;AAC7E,MAAM,oBAAoB,GAAG,CAAC,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,CAAU,CAAC;AAKzF,8EAA8E;AAC9E,UAAU;AACV,8EAA8E;AAE9E;;;;;GAKG;AACH,SAAS,KAAK,CAAC,QAAgB;IAC7B,OAAO,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;AAC9C,CAAC;AAED;;;;;;;GAOG;AACH,SAAS,SAAS,CAAC,IAAY;IAC7B,OAAO,IAAI,CAAC,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC;AAClE,CAAC;AAED;;;;;GAKG;AACH,SAAS,WAAW,CAAC,GAA0B;IAC7C,QAAQ,GAAmB,EAAE,CAAC;QAC5B,KAAK,MAAM;YACT,OAAO,KAAK,CAAC;QACf,KAAK,MAAM;YACT,OAAO,KAAK,CAAC;QACf,KAAK,MAAM;YACT,OAAO,KAAK,CAAC;QACf,KAAK,OAAO;YACV,OAAO,MAAM,CAAC;QAChB,KAAK,OAAO,CAAC;QACb,KAAK,MAAM;YACT,OAAO,MAAM,CAAC;QAChB;YACE,OAAO,KAAK,CAAC;IACjB,CAAC;AACH,CAAC;AAED;;;;;;;;;;GAUG;AACH,SAAS,UAAU,CAAC,GAAW;IAC7B,IAAI,CAAC;QACH,OAAO,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC;IAClD,CAAC;IAAC,MAAM,CAAC;QACP,qEAAqE;QACrE,OAAO,GAAG,CAAC;IACb,CAAC;AACH,CAAC;AAED,8EAA8E;AAC9E,aAAa;AACb,8EAA8E;AAE9E;;;;;;;;;;;;;;;;;;;;;GAqBG;AACH,MAAM,OAAO,UAAU;IAAvB;QACE,kBAAkB;QACT,wBAAmB,GAAa,CAAC,GAAG,oBAAoB,CAAC,CAAC;IAqGrE,CAAC;IAnGC,4EAA4E;IAC5E,UAAU;IACV,4EAA4E;IAE5E,kBAAkB;IAClB,OAAO,CAAC,MAAuB;QAC7B,sEAAsE;QACtE,uDAAuD;QACvD,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,OAAO,KAAK,CAAC;QACf,CAAC;QACD,OAAQ,oBAA0C,CAAC,QAAQ,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;IAC7E,CAAC;IAED,4EAA4E;IAC5E,OAAO;IACP,4EAA4E;IAE5E,kBAAkB;IAClB,KAAK,CAAC,IAAI,CAAC,MAAuB,EAAE,QAAsB;QACxD,IAAI,GAAW,CAAC;QAChB,IAAI,YAAgC,CAAC;QACrC,IAAI,GAAW,CAAC;QAEhB,IAAI,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC5B,sEAAsE;YACtE,GAAG,GAAG,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;YAC9B,GAAG,GAAG,MAAM,CAAC;QACf,CAAC;aAAM,CAAC;YACN,4DAA4D;YAC5D,YAAY,GAAG,YAAY,CAAC,MAAM,CAAC,CAAC;YACpC,GAAG,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC;YACpB,MAAM,KAAK,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC;YAC9C,GAAG,GAAG,KAAK,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;QAC/B,CAAC;QAED,kCAAkC;QAClC,MAAM,OAAO,GAAG,IAAI,CAAC,UAAU,CAAC,GAAG,EAAE,GAAmB,CAAC,CAAC;QAE1D,8BAA8B;QAC9B,MAAM,IAAI,GAAqB;YAC7B,SAAS,EAAE,SAAS,CAAC,GAAG,CAAC;YACzB,MAAM,EAAE,WAAW,CAAC,GAAG,CAAC;YACxB,GAAG,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SAClD,CAAC;QAEF,OAAO;YACL,OAAO;YACP,QAAQ,EAAE,IAAI;YACd,MAAM,EAAE,WAAW,CAAC,GAAG,CAAC;SACzB,CAAC;IACJ,CAAC;IAED,4EAA4E;IAC5E,kBAAkB;IAClB,4EAA4E;IAE5E;;;;;;;;;OASG;IACK,UAAU,CAAC,GAAW,EAAE,GAA0B;QACxD,QAAQ,GAAmB,EAAE,CAAC;YAC5B,KAAK,OAAO;gBACV,OAAO,UAAU,CAAC,GAAG,CAAC,CAAC;YAEzB,KAAK,OAAO,CAAC;YACb,KAAK,MAAM;gBACT,OAAO,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;YAE/B;gBACE,mDAAmD;gBACnD,OAAO,GAAG,CAAC;QACf,CAAC;IACH,CAAC;IAED;;;;;;;;OAQG;IACK,WAAW,CAAC,GAAW;QAC7B,IAAI,CAAC;YACH,MAAM,MAAM,GAAY,SAAS,CAAC,GAAG,CAAC,CAAC;YACvC,OAAO,aAAa,CAAC,MAAM,CAAC,CAAC;QAC/B,CAAC;QAAC,MAAM,CAAC;YACP,OAAO,GAAG,CAAC;QACb,CAAC;IACH,CAAC;CACF"}