@mcp-graph-workflow/mcp-graph 3.0.0

package/dist/core/integrations/serena-reader.js

@@ -0,0 +1,68 @@
+import path from "node:path";
+import { readdir, readFile } from "node:fs/promises";
+import { logger } from "../utils/logger.js";
+const SERENA_MEMORIES_DIR = ".serena/memories";
+/**
+ * Recursively collect all .md files under a directory, returning paths relative to root.
+ */
+async function collectMdFiles(dir, root) {
+    const entries = await readdir(dir, { withFileTypes: true });
+    const results = [];
+    for (const entry of entries) {
+        const fullPath = path.join(dir, entry.name);
+        if (entry.isDirectory()) {
+            const nested = await collectMdFiles(fullPath, root);
+            results.push(...nested);
+        }
+        else if (entry.isFile() && entry.name.endsWith(".md")) {
+            const relative = path.relative(root, fullPath);
+            results.push(relative.replace(/\.md$/, ""));
+        }
+    }
+    return results;
+}
+/**
+ * List all Serena memory files from the project directory (supports nested dirs).
+ */
+export async function listSerenaMemories(basePath) {
+    try {
+        const dir = path.join(basePath, SERENA_MEMORIES_DIR);
+        return await collectMdFiles(dir, dir);
+    }
+    catch {
+        logger.info("No Serena memories directory found", { basePath });
+        return [];
+    }
+}
+/**
+ * Read a specific Serena memory file.
+ */
+export async function readSerenaMemory(basePath, name) {
+    try {
+        const filePath = path.join(basePath, SERENA_MEMORIES_DIR, `${name}.md`);
+        const content = await readFile(filePath, "utf-8");
+        return {
+            name,
+            content,
+            sizeBytes: Buffer.byteLength(content, "utf-8"),
+        };
+    }
+    catch {
+        logger.info("Serena memory not found", { name });
+        return null;
+    }
+}
+/**
+ * Read all Serena memories at once.
+ */
+export async function readAllSerenaMemories(basePath) {
+    const names = await listSerenaMemories(basePath);
+    const memories = [];
+    for (const name of names) {
+        const memory = await readSerenaMemory(basePath, name);
+        if (memory)
+            memories.push(memory);
+    }
+    return memories;
+}
+//# sourceMappingURL=serena-reader.js.map

package/dist/core/integrations/serena-reader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"serena-reader.js","sourceRoot":"","sources":["../../../src/core/integrations/serena-reader.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AACrD,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAQ5C,MAAM,mBAAmB,GAAG,kBAAkB,CAAC;AAE/C;;GAEG;AACH,KAAK,UAAU,cAAc,CAAC,GAAW,EAAE,IAAY;IACrD,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,GAAG,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;IAC5D,MAAM,OAAO,GAAa,EAAE,CAAC;IAE7B,KAAK,MAAM,KAAK,IAAI,OAAO,EAAE,CAAC;QAC5B,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,CAAC,CAAC;QAC5C,IAAI,KAAK,CAAC,WAAW,EAAE,EAAE,CAAC;YACxB,MAAM,MAAM,GAAG,MAAM,cAAc,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC;YACpD,OAAO,CAAC,IAAI,CAAC,GAAG,MAAM,CAAC,CAAC;QAC1B,CAAC;aAAM,IAAI,KAAK,CAAC,MAAM,EAAE,IAAI,KAAK,CAAC,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC;YACxD,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAC;YAC/C,OAAO,CAAC,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,CAAC;QAC9C,CAAC;IACH,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IACvD,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,mBAAmB,CAAC,CAAC;QACrD,OAAO,MAAM,cAAc,CAAC,GAAG,EAAE,GAAG,CAAC,CAAC;IACxC,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,CAAC,IAAI,CAAC,oCAAoC,EAAE,EAAE,QAAQ,EAAE,CAAC,CAAC;QAChE,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,gBAAgB,CAAC,QAAgB,EAAE,IAAY;IACnE,IAAI,CAAC;QACH,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,mBAAmB,EAAE,GAAG,IAAI,KAAK,CAAC,CAAC;QACxE,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;QAClD,OAAO;YACL,IAAI;YACJ,OAAO;YACP,SAAS,EAAE,MAAM,CAAC,UAAU,CAAC,OAAO,EAAE,OAAO,CAAC;SAC/C,CAAC;IACJ,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,CAAC,IAAI,CAAC,yBAAyB,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC;QACjD,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,qBAAqB,CAAC,QAAgB;IAC1D,MAAM,KAAK,GAAG,MAAM,kBAAkB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,QAAQ,GAAmB,EAAE,CAAC;IAEpC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,MAAM,GAAG,MAAM,gBAAgB,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC;QACtD,IAAI,MAAM;YAAE,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IACpC,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/core/integrations/tool-status.d.ts

@@ -0,0 +1,18 @@
+export interface ToolInfo {
+    installed: boolean;
+    running: boolean;
+    url?: string;
+    details?: Record<string, unknown>;
+}
+export interface IntegrationsStatus {
+    gitnexus: ToolInfo;
+    serena: ToolInfo & {
+        memories: string[];
+    };
+    playwright: ToolInfo;
+}
+/**
+ * Detect status of all ecosystem tools.
+ */
+export declare function getIntegrationsStatus(basePath: string): Promise<IntegrationsStatus>;
+//# sourceMappingURL=tool-status.d.ts.map

package/dist/core/integrations/tool-status.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"tool-status.d.ts","sourceRoot":"","sources":["../../../src/core/integrations/tool-status.ts"],"names":[],"mappings":"AAGA,MAAM,WAAW,QAAQ;IACvB,SAAS,EAAE,OAAO,CAAC;IACnB,OAAO,EAAE,OAAO,CAAC;IACjB,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACnC;AAED,MAAM,WAAW,kBAAkB;IACjC,QAAQ,EAAE,QAAQ,CAAC;IACnB,MAAM,EAAE,QAAQ,GAAG;QAAE,QAAQ,EAAE,MAAM,EAAE,CAAA;KAAE,CAAC;IAC1C,UAAU,EAAE,QAAQ,CAAC;CACtB;AAiED;;GAEG;AACH,wBAAsB,qBAAqB,CAAC,QAAQ,EAAE,MAAM,GAAG,OAAO,CAAC,kBAAkB,CAAC,CA+BzF"}

package/dist/core/integrations/tool-status.js

@@ -0,0 +1,97 @@
+import { logger } from "../utils/logger.js";
+import { whichCommand } from "../utils/platform.js";
+/**
+ * Detect whether a local HTTP service is running by attempting a fetch.
+ */
+async function probeHttp(url, timeoutMs = 2000) {
+    try {
+        const controller = new AbortController();
+        const timer = setTimeout(() => controller.abort(), timeoutMs);
+        const res = await fetch(url, { signal: controller.signal });
+        clearTimeout(timer);
+        return res.ok || res.status < 500;
+    }
+    catch (err) {
+        logger.debug("probe:http:fail", { url, error: err instanceof Error ? err.message : String(err) });
+        return false;
+    }
+}
+/**
+ * Check if a command-line tool is installed by looking for it in PATH.
+ */
+async function isCommandInstalled(command) {
+    try {
+        const { execFile } = await import("node:child_process");
+        const { promisify } = await import("node:util");
+        const exec = promisify(execFile);
+        await exec(whichCommand(), [command]);
+        return true;
+    }
+    catch (err) {
+        logger.debug("probe:command:fail", { command, error: err instanceof Error ? err.message : String(err) });
+        return false;
+    }
+}
+/**
+ * Read Serena memory file names from .serena/memories/ directory.
+ */
+async function readSerenaMemories(basePath) {
+    try {
+        const path = await import("node:path");
+        const { readdir } = await import("node:fs/promises");
+        const memoriesDir = path.join(basePath, ".serena", "memories");
+        const files = await readdir(memoriesDir);
+        return files.filter((f) => f.endsWith(".md")).map((f) => f.replace(/\.md$/, ""));
+    }
+    catch (err) {
+        logger.debug("serena:memories:fail", { error: err instanceof Error ? err.message : String(err) });
+        return [];
+    }
+}
+/**
+ * Check if Serena is configured for this project.
+ */
+async function isSerenaConfigured(basePath) {
+    try {
+        const path = await import("node:path");
+        const { access } = await import("node:fs/promises");
+        await access(path.join(basePath, ".serena"));
+        return true;
+    }
+    catch (err) {
+        logger.debug("serena:config:fail", { basePath, error: err instanceof Error ? err.message : String(err) });
+        return false;
+    }
+}
+/**
+ * Detect status of all ecosystem tools.
+ */
+export async function getIntegrationsStatus(basePath) {
+    logger.info("Checking integrations status", { basePath });
+    const GITNEXUS_PORT = parseInt(process.env.GITNEXUS_PORT || "3737", 10);
+    const gitnexusUrl = `http://localhost:${GITNEXUS_PORT}`;
+    const [gitnexusInstalled, gitnexusRunning, serenaConfigured, serenaMemories, playwrightInstalled] = await Promise.all([
+        isCommandInstalled("gitnexus"),
+        probeHttp(gitnexusUrl),
+        isSerenaConfigured(basePath),
+        readSerenaMemories(basePath),
+        isCommandInstalled("npx").then(() => true).catch(() => false),
+    ]);
+    return {
+        gitnexus: {
+            installed: gitnexusInstalled,
+            running: gitnexusRunning,
+            ...(gitnexusRunning ? { url: gitnexusUrl } : {}),
+        },
+        serena: {
+            installed: serenaConfigured,
+            running: serenaConfigured,
+            memories: serenaMemories,
+        },
+        playwright: {
+            installed: playwrightInstalled,
+            running: false,
+        },
+    };
+}
+//# sourceMappingURL=tool-status.js.map

package/dist/core/integrations/tool-status.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"tool-status.js","sourceRoot":"","sources":["../../../src/core/integrations/tool-status.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAC5C,OAAO,EAAE,YAAY,EAAE,MAAM,sBAAsB,CAAC;AAepD;;GAEG;AACH,KAAK,UAAU,SAAS,CAAC,GAAW,EAAE,YAAoB,IAAI;IAC5D,IAAI,CAAC;QACH,MAAM,UAAU,GAAG,IAAI,eAAe,EAAE,CAAC;QACzC,MAAM,KAAK,GAAG,UAAU,CAAC,GAAG,EAAE,CAAC,UAAU,CAAC,KAAK,EAAE,EAAE,SAAS,CAAC,CAAC;QAC9D,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,GAAG,EAAE,EAAE,MAAM,EAAE,UAAU,CAAC,MAAM,EAAE,CAAC,CAAC;QAC5D,YAAY,CAAC,KAAK,CAAC,CAAC;QACpB,OAAO,GAAG,CAAC,EAAE,IAAI,GAAG,CAAC,MAAM,GAAG,GAAG,CAAC;IACpC,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,MAAM,CAAC,KAAK,CAAC,iBAAiB,EAAE,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QAClG,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,kBAAkB,CAAC,OAAe;IAC/C,IAAI,CAAC;QACH,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,MAAM,CAAC,oBAAoB,CAAC,CAAC;QACxD,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,WAAW,CAAC,CAAC;QAChD,MAAM,IAAI,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;QACjC,MAAM,IAAI,CAAC,YAAY,EAAE,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC;QACtC,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,MAAM,CAAC,KAAK,CAAC,oBAAoB,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACzG,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IAChD,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,MAAM,MAAM,CAAC,WAAW,CAAC,CAAC;QACvC,MAAM,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,kBAAkB,CAAC,CAAC;QACrD,MAAM,WAAW,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,SAAS,EAAE,UAAU,CAAC,CAAC;QAC/D,MAAM,KAAK,GAAG,MAAM,OAAO,CAAC,WAAW,CAAC,CAAC;QACzC,OAAO,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,CAAC;IACnF,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,MAAM,CAAC,KAAK,CAAC,sBAAsB,EAAE,EAAE,KAAK,EAAE,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QAClG,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IAChD,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,MAAM,MAAM,CAAC,WAAW,CAAC,CAAC;QACvC,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,MAAM,CAAC,kBAAkB,CAAC,CAAC;QACpD,MAAM,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,SAAS,CAAC,CAAC,CAAC;QAC7C,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,MAAM,CAAC,KAAK,CAAC,oBAAoB,EAAE,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QAC1G,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,qBAAqB,CAAC,QAAgB;IAC1D,MAAM,CAAC,IAAI,CAAC,8BAA8B,EAAE,EAAE,QAAQ,EAAE,CAAC,CAAC;IAE1D,MAAM,aAAa,GAAG,QAAQ,CAAC,OAAO,CAAC,GAAG,CAAC,aAAa,IAAI,MAAM,EAAE,EAAE,CAAC,CAAC;IACxE,MAAM,WAAW,GAAG,oBAAoB,aAAa,EAAE,CAAC;IAExD,MAAM,CAAC,iBAAiB,EAAE,eAAe,EAAE,gBAAgB,EAAE,cAAc,EAAE,mBAAmB,CAAC,GAC/F,MAAM,OAAO,CAAC,GAAG,CAAC;QAChB,kBAAkB,CAAC,UAAU,CAAC;QAC9B,SAAS,CAAC,WAAW,CAAC;QACtB,kBAAkB,CAAC,QAAQ,CAAC;QAC5B,kBAAkB,CAAC,QAAQ,CAAC;QAC5B,kBAAkB,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC;KAC9D,CAAC,CAAC;IAEL,OAAO;QACL,QAAQ,EAAE;YACR,SAAS,EAAE,iBAAiB;YAC5B,OAAO,EAAE,eAAe;YACxB,GAAG,CAAC,eAAe,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,WAAW,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SACjD;QACD,MAAM,EAAE;YACN,SAAS,EAAE,gBAAgB;YAC3B,OAAO,EAAE,gBAAgB;YACzB,QAAQ,EAAE,cAAc;SACzB;QACD,UAAU,EAAE;YACV,SAAS,EAAE,mBAAmB;YAC9B,OAAO,EAAE,KAAK;SACf;KACF,CAAC;AACJ,CAAC"}

package/dist/core/parser/classify.d.ts

@@ -0,0 +1,31 @@
+/**
+ * Stage 3: Heuristic classification.
+ * Classify sections and bullet items into semantic types.
+ */
+import type { NodeType } from "../graph/graph-types.js";
+export type BlockType = NodeType | "unknown";
+export interface ClassifiedBlock {
+    type: BlockType;
+    title: string;
+    description: string;
+    items: ClassifiedItem[];
+    startLine: number;
+    endLine: number;
+    confidence: number;
+}
+export interface ClassifiedItem {
+    type: BlockType;
+    text: string;
+    line: number;
+    confidence: number;
+}
+export declare function classifyText(text: string): {
+    type: BlockType;
+    confidence: number;
+};
+export declare function classifySectionTitle(title: string, level: number): {
+    type: BlockType;
+    confidence: number;
+};
+export declare function classifySection(title: string, body: string, level: number, startLine: number, endLine: number): ClassifiedBlock;
+//# sourceMappingURL=classify.d.ts.map

package/dist/core/parser/classify.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"classify.d.ts","sourceRoot":"","sources":["../../../src/core/parser/classify.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,KAAK,EAAE,QAAQ,EAAE,MAAM,yBAAyB,CAAC;AAExD,MAAM,MAAM,SAAS,GAAG,QAAQ,GAAG,SAAS,CAAC;AAE7C,MAAM,WAAW,eAAe;IAC9B,IAAI,EAAE,SAAS,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,WAAW,EAAE,MAAM,CAAC;IACpB,KAAK,EAAE,cAAc,EAAE,CAAC;IACxB,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;IAChB,UAAU,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE,SAAS,CAAC;IAChB,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,UAAU,EAAE,MAAM,CAAC;CACpB;AAwCD,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG;IAAE,IAAI,EAAE,SAAS,CAAC;IAAC,UAAU,EAAE,MAAM,CAAA;CAAE,CAQlF;AAED,wBAAgB,oBAAoB,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,GAAG;IAAE,IAAI,EAAE,SAAS,CAAC;IAAC,UAAU,EAAE,MAAM,CAAA;CAAE,CAY1G;AAgDD,wBAAgB,eAAe,CAC7B,KAAK,EAAE,MAAM,EACb,IAAI,EAAE,MAAM,EACZ,KAAK,EAAE,MAAM,EACb,SAAS,EAAE,MAAM,EACjB,OAAO,EAAE,MAAM,GACd,eAAe,CA4BjB"}

package/dist/core/parser/classify.js

@@ -0,0 +1,128 @@
+/**
+ * Stage 3: Heuristic classification.
+ * Classify sections and bullet items into semantic types.
+ */
+// --- Heuristic keyword patterns (Portuguese + English) ---
+const REQUIREMENT_PATTERNS = [
+    /\bdeve\b/i, /\bprecisa\b/i, /\bnecessário\b/i, /\bmust\b/i,
+    /\bshould\b/i, /\brequired\b/i, /\brequisito/i, /\brequirement/i,
+];
+const CONSTRAINT_PATTERNS = [
+    /\bnão deve\b/i, /\bsem\b/i, /\brestrição/i, /\bnão depender/i,
+    /\bnão exigir/i, /\bconstraint/i, /\bwithout\b/i, /\bnot allowed/i,
+    /\bfora do escopo/i,
+];
+const TASK_PATTERNS = [
+    /\bimplementar\b/i, /\bcriar\b/i, /\badicionar\b/i, /\bdefinir\b/i,
+    /\bconstruir\b/i, /\bconfigurar\b/i, /\binstalar\b/i, /\bdesenvolver\b/i,
+    /\bimplement\b/i, /\bcreate\b/i, /\bbuild\b/i, /\bset up\b/i,
+    /\bdesign\b/i,
+];
+const ACCEPTANCE_PATTERNS = [
+    /\baceite\b/i, /\bcritério/i, /\bdone\b/i, /\bacceptance/i,
+    /\bcriterion/i, /\bcriteria/i, /\bdefinition of done/i,
+];
+const RISK_PATTERNS = [
+    /\brisco\b/i, /\brisk\b/i, /\bmitigação/i, /\bmitigation/i,
+];
+const EPIC_TITLE_PATTERNS = [
+    /\bepic\b/i, /\bvisão\b/i, /\bvision\b/i, /\bobjetivo principal/i,
+    /\bproduto\b/i, /\bprojeto\b/i,
+];
+function matchesAny(text, patterns) {
+    return patterns.some((p) => p.test(text));
+}
+export function classifyText(text) {
+    // Order matters: more specific checks first
+    if (matchesAny(text, CONSTRAINT_PATTERNS))
+        return { type: "constraint", confidence: 0.8 };
+    if (matchesAny(text, ACCEPTANCE_PATTERNS))
+        return { type: "acceptance_criteria", confidence: 0.8 };
+    if (matchesAny(text, RISK_PATTERNS))
+        return { type: "risk", confidence: 0.7 };
+    if (matchesAny(text, TASK_PATTERNS))
+        return { type: "task", confidence: 0.7 };
+    if (matchesAny(text, REQUIREMENT_PATTERNS))
+        return { type: "requirement", confidence: 0.7 };
+    return { type: "unknown", confidence: 0.3 };
+}
+export function classifySectionTitle(title, level) {
+    const lower = title.toLowerCase();
+    if (matchesAny(lower, ACCEPTANCE_PATTERNS))
+        return { type: "acceptance_criteria", confidence: 0.9 };
+    if (matchesAny(lower, RISK_PATTERNS))
+        return { type: "risk", confidence: 0.85 };
+    if (matchesAny(lower, CONSTRAINT_PATTERNS))
+        return { type: "constraint", confidence: 0.85 };
+    if (/\brequisito/i.test(lower) || /\brequirement/i.test(lower))
+        return { type: "requirement", confidence: 0.9 };
+    if (level === 1 || matchesAny(lower, EPIC_TITLE_PATTERNS))
+        return { type: "epic", confidence: 0.8 };
+    if (/\btask\b/i.test(lower) || /\bentrega/i.test(lower))
+        return { type: "task", confidence: 0.85 };
+    return { type: "unknown", confidence: 0.3 };
+}
+function parseBulletItems(body, startLine) {
+    const items = [];
+    const lines = body.split("\n");
+    for (let i = 0; i < lines.length; i++) {
+        const line = lines[i];
+        const bulletMatch = line.match(/^\s*[-]\s+(.+)$/);
+        if (!bulletMatch)
+            continue;
+        const text = bulletMatch[1].trim();
+        const { type, confidence } = classifyText(text);
+        items.push({
+            type,
+            text,
+            line: startLine + i,
+            confidence,
+        });
+    }
+    return items;
+}
+function parseNumberedItems(body, startLine) {
+    const items = [];
+    const lines = body.split("\n");
+    for (let i = 0; i < lines.length; i++) {
+        const line = lines[i];
+        const numMatch = line.match(/^\s*\d+[.)]\s+(.+)$/);
+        if (!numMatch)
+            continue;
+        const text = numMatch[1].trim();
+        const { type, confidence } = classifyText(text);
+        items.push({
+            type,
+            text,
+            line: startLine + i,
+            confidence,
+        });
+    }
+    return items;
+}
+export function classifySection(title, body, level, startLine, endLine) {
+    const { type: sectionType, confidence: sectionConf } = classifySectionTitle(title, level);
+    const bulletItems = parseBulletItems(body, startLine);
+    const numberedItems = parseNumberedItems(body, startLine);
+    const items = [...bulletItems, ...numberedItems];
+    // If the section is a generic "unknown" but its items are mostly tasks, promote it
+    let finalType = sectionType;
+    let finalConf = sectionConf;
+    if (sectionType === "unknown" && items.length > 0) {
+        const taskCount = items.filter((it) => it.type === "task").length;
+        if (taskCount > items.length / 2) {
+            finalType = "task";
+            finalConf = 0.6;
+        }
+    }
+    return {
+        type: finalType,
+        title,
+        description: body,
+        items,
+        startLine,
+        endLine,
+        confidence: finalConf,
+    };
+}
+//# sourceMappingURL=classify.js.map

package/dist/core/parser/classify.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"classify.js","sourceRoot":"","sources":["../../../src/core/parser/classify.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAuBH,4DAA4D;AAE5D,MAAM,oBAAoB,GAAG;IAC3B,WAAW,EAAE,cAAc,EAAE,iBAAiB,EAAE,WAAW;IAC3D,aAAa,EAAE,eAAe,EAAE,cAAc,EAAE,gBAAgB;CACjE,CAAC;AAEF,MAAM,mBAAmB,GAAG;IAC1B,eAAe,EAAE,UAAU,EAAE,cAAc,EAAE,iBAAiB;IAC9D,eAAe,EAAE,eAAe,EAAE,cAAc,EAAE,gBAAgB;IAClE,mBAAmB;CACpB,CAAC;AAEF,MAAM,aAAa,GAAG;IACpB,kBAAkB,EAAE,YAAY,EAAE,gBAAgB,EAAE,cAAc;IAClE,gBAAgB,EAAE,iBAAiB,EAAE,eAAe,EAAE,kBAAkB;IACxE,gBAAgB,EAAE,aAAa,EAAE,YAAY,EAAE,aAAa;IAC5D,aAAa;CACd,CAAC;AAEF,MAAM,mBAAmB,GAAG;IAC1B,aAAa,EAAE,aAAa,EAAE,WAAW,EAAE,eAAe;IAC1D,cAAc,EAAE,aAAa,EAAE,uBAAuB;CACvD,CAAC;AAEF,MAAM,aAAa,GAAG;IACpB,YAAY,EAAE,WAAW,EAAE,cAAc,EAAE,eAAe;CAC3D,CAAC;AAEF,MAAM,mBAAmB,GAAG;IAC1B,WAAW,EAAE,YAAY,EAAE,aAAa,EAAE,uBAAuB;IACjE,cAAc,EAAE,cAAc;CAC/B,CAAC;AAEF,SAAS,UAAU,CAAC,IAAY,EAAE,QAAkB;IAClD,OAAO,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC;AAC5C,CAAC;AAED,MAAM,UAAU,YAAY,CAAC,IAAY;IACvC,4CAA4C;IAC5C,IAAI,UAAU,CAAC,IAAI,EAAE,mBAAmB,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IAC1F,IAAI,UAAU,CAAC,IAAI,EAAE,mBAAmB,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,qBAAqB,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IACnG,IAAI,UAAU,CAAC,IAAI,EAAE,aAAa,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IAC9E,IAAI,UAAU,CAAC,IAAI,EAAE,aAAa,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IAC9E,IAAI,UAAU,CAAC,IAAI,EAAE,oBAAoB,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,aAAa,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IAC5F,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;AAC9C,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAC,KAAa,EAAE,KAAa;IAC/D,MAAM,KAAK,GAAG,KAAK,CAAC,WAAW,EAAE,CAAC;IAElC,IAAI,UAAU,CAAC,KAAK,EAAE,mBAAmB,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,qBAAqB,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IACpG,IAAI,UAAU,CAAC,KAAK,EAAE,aAAa,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC;IAChF,IAAI,UAAU,CAAC,KAAK,EAAE,mBAAmB,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC;IAC5F,IAAI,cAAc,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,gBAAgB,CAAC,IAAI,CAAC,KAAK,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,aAAa,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IAEhH,IAAI,KAAK,KAAK,CAAC,IAAI,UAAU,CAAC,KAAK,EAAE,mBAAmB,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;IACpG,IAAI,WAAW,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,YAAY,CAAC,IAAI,CAAC,KAAK,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC;IAEnG,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;AAC9C,CAAC;AAED,SAAS,gBAAgB,CAAC,IAAY,EAAE,SAAiB;IACvD,MAAM,KAAK,GAAqB,EAAE,CAAC;IACnC,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAE/B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACtB,MAAM,WAAW,GAAG,IAAI,CAAC,KAAK,CAAC,iBAAiB,CAAC,CAAC;QAClD,IAAI,CAAC,WAAW;YAAE,SAAS;QAE3B,MAAM,IAAI,GAAG,WAAW,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACnC,MAAM,EAAE,IAAI,EAAE,UAAU,EAAE,GAAG,YAAY,CAAC,IAAI,CAAC,CAAC;QAEhD,KAAK,CAAC,IAAI,CAAC;YACT,IAAI;YACJ,IAAI;YACJ,IAAI,EAAE,SAAS,GAAG,CAAC;YACnB,UAAU;SACX,CAAC,CAAC;IACL,CAAC;IAED,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,kBAAkB,CAAC,IAAY,EAAE,SAAiB;IACzD,MAAM,KAAK,GAAqB,EAAE,CAAC;IACnC,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAE/B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACtB,MAAM,QAAQ,GAAG,IAAI,CAAC,KAAK,CAAC,qBAAqB,CAAC,CAAC;QACnD,IAAI,CAAC,QAAQ;YAAE,SAAS;QAExB,MAAM,IAAI,GAAG,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAChC,MAAM,EAAE,IAAI,EAAE,UAAU,EAAE,GAAG,YAAY,CAAC,IAAI,CAAC,CAAC;QAEhD,KAAK,CAAC,IAAI,CAAC;YACT,IAAI;YACJ,IAAI;YACJ,IAAI,EAAE,SAAS,GAAG,CAAC;YACnB,UAAU;SACX,CAAC,CAAC;IACL,CAAC;IAED,OAAO,KAAK,CAAC;AACf,CAAC;AAED,MAAM,UAAU,eAAe,CAC7B,KAAa,EACb,IAAY,EACZ,KAAa,EACb,SAAiB,EACjB,OAAe;IAEf,MAAM,EAAE,IAAI,EAAE,WAAW,EAAE,UAAU,EAAE,WAAW,EAAE,GAAG,oBAAoB,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC;IAE1F,MAAM,WAAW,GAAG,gBAAgB,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;IACtD,MAAM,aAAa,GAAG,kBAAkB,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;IAC1D,MAAM,KAAK,GAAG,CAAC,GAAG,WAAW,EAAE,GAAG,aAAa,CAAC,CAAC;IAEjD,mFAAmF;IACnF,IAAI,SAAS,GAAG,WAAW,CAAC;IAC5B,IAAI,SAAS,GAAG,WAAW,CAAC;IAE5B,IAAI,WAAW,KAAK,SAAS,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAClD,MAAM,SAAS,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,EAAE,EAAE,EAAE,CAAC,EAAE,CAAC,IAAI,KAAK,MAAM,CAAC,CAAC,MAAM,CAAC;QAClE,IAAI,SAAS,GAAG,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACjC,SAAS,GAAG,MAAM,CAAC;YACnB,SAAS,GAAG,GAAG,CAAC;QAClB,CAAC;IACH,CAAC;IAED,OAAO;QACL,IAAI,EAAE,SAAS;QACf,KAAK;QACL,WAAW,EAAE,IAAI;QACjB,KAAK;QACL,SAAS;QACT,OAAO;QACP,UAAU,EAAE,SAAS;KACtB,CAAC;AACJ,CAAC"}

package/dist/core/parser/extract.d.ts

@@ -0,0 +1,21 @@
+/**
+ * Stage 4: Entity extraction.
+ * Combine normalize → segment → classify to produce structured extraction results.
+ */
+import type { ClassifiedBlock } from "./classify.js";
+export interface ExtractionResult {
+    blocks: ClassifiedBlock[];
+    summary: {
+        totalSections: number;
+        epics: number;
+        tasks: number;
+        subtasks: number;
+        requirements: number;
+        constraints: number;
+        acceptanceCriteria: number;
+        risks: number;
+        unknown: number;
+    };
+}
+export declare function extractEntities(rawText: string): ExtractionResult;
+//# sourceMappingURL=extract.d.ts.map

package/dist/core/parser/extract.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"extract.d.ts","sourceRoot":"","sources":["../../../src/core/parser/extract.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAKH,OAAO,KAAK,EAAE,eAAe,EAA6B,MAAM,eAAe,CAAC;AAGhF,MAAM,WAAW,gBAAgB;IAC/B,MAAM,EAAE,eAAe,EAAE,CAAC;IAC1B,OAAO,EAAE;QACP,aAAa,EAAE,MAAM,CAAC;QACtB,KAAK,EAAE,MAAM,CAAC;QACd,KAAK,EAAE,MAAM,CAAC;QACd,QAAQ,EAAE,MAAM,CAAC;QACjB,YAAY,EAAE,MAAM,CAAC;QACrB,WAAW,EAAE,MAAM,CAAC;QACpB,kBAAkB,EAAE,MAAM,CAAC;QAC3B,KAAK,EAAE,MAAM,CAAC;QACd,OAAO,EAAE,MAAM,CAAC;KACjB,CAAC;CACH;AAQD,wBAAgB,eAAe,CAAC,OAAO,EAAE,MAAM,GAAG,gBAAgB,CA2DjE"}

package/dist/core/parser/extract.js

@@ -0,0 +1,66 @@
+/**
+ * Stage 4: Entity extraction.
+ * Combine normalize → segment → classify to produce structured extraction results.
+ */
+import { normalize } from "./normalize.js";
+import { segment } from "./segment.js";
+import { classifySection, classifyText } from "./classify.js";
+import { logger } from "../utils/logger.js";
+function countByType(blocks, items, type) {
+    const blockCount = blocks.filter((b) => b.type === type).length;
+    const itemCount = items.filter((i) => i.type === type).length;
+    return blockCount + itemCount;
+}
+export function extractEntities(rawText) {
+    logger.info(`Extracting entities from ${rawText.length} chars`);
+    const normalized = normalize(rawText);
+    const sections = segment(normalized);
+    logger.info(`Segmented into ${sections.length} sections`);
+    const blocks = sections.map((sec) => classifySection(sec.title, sec.body, sec.level, sec.startLine, sec.endLine));
+    // Promote items inside task-sections to subtask if they're generic or tasks
+    for (const block of blocks) {
+        if (block.type === "task" || block.type === "epic") {
+            for (const item of block.items) {
+                if (item.type === "unknown" || item.type === "task") {
+                    // Items inside a task/epic section are subtasks
+                    if (block.type === "task") {
+                        item.type = "subtask";
+                        item.confidence = Math.max(item.confidence, 0.6);
+                    }
+                }
+            }
+        }
+    }
+    // Also classify numbered items inside sections that look like task lists
+    // (e.g., "Entregas" sections with numbered action items)
+    for (const block of blocks) {
+        if (block.type === "unknown") {
+            for (const item of block.items) {
+                if (item.type === "unknown") {
+                    const reclassified = classifyText(item.text);
+                    if (reclassified.type !== "unknown") {
+                        item.type = reclassified.type;
+                        item.confidence = reclassified.confidence;
+                    }
+                }
+            }
+        }
+    }
+    const allItems = blocks.flatMap((b) => b.items);
+    logger.info(`Extraction complete: ${blocks.length} blocks, ${allItems.length} items`);
+    return {
+        blocks,
+        summary: {
+            totalSections: blocks.length,
+            epics: countByType(blocks, allItems, "epic"),
+            tasks: countByType(blocks, allItems, "task"),
+            subtasks: countByType(blocks, allItems, "subtask"),
+            requirements: countByType(blocks, allItems, "requirement"),
+            constraints: countByType(blocks, allItems, "constraint"),
+            acceptanceCriteria: countByType(blocks, allItems, "acceptance_criteria"),
+            risks: countByType(blocks, allItems, "risk"),
+            unknown: countByType(blocks, allItems, "unknown"),
+        },
+    };
+}
+//# sourceMappingURL=extract.js.map

package/dist/core/parser/extract.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extract.js","sourceRoot":"","sources":["../../../src/core/parser/extract.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,OAAO,EAAE,SAAS,EAAE,MAAM,gBAAgB,CAAC;AAC3C,OAAO,EAAE,OAAO,EAAE,MAAM,cAAc,CAAC;AACvC,OAAO,EAAE,eAAe,EAAE,YAAY,EAAE,MAAM,eAAe,CAAC;AAE9D,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAiB5C,SAAS,WAAW,CAAC,MAAyB,EAAE,KAAuB,EAAE,IAAe;IACtF,MAAM,UAAU,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,IAAI,CAAC,CAAC,MAAM,CAAC;IAChE,MAAM,SAAS,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,IAAI,CAAC,CAAC,MAAM,CAAC;IAC9D,OAAO,UAAU,GAAG,SAAS,CAAC;AAChC,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,OAAe;IAC7C,MAAM,CAAC,IAAI,CAAC,4BAA4B,OAAO,CAAC,MAAM,QAAQ,CAAC,CAAC;IAChE,MAAM,UAAU,GAAG,SAAS,CAAC,OAAO,CAAC,CAAC;IACtC,MAAM,QAAQ,GAAG,OAAO,CAAC,UAAU,CAAC,CAAC;IACrC,MAAM,CAAC,IAAI,CAAC,kBAAkB,QAAQ,CAAC,MAAM,WAAW,CAAC,CAAC;IAE1D,MAAM,MAAM,GAAsB,QAAQ,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CACrD,eAAe,CAAC,GAAG,CAAC,KAAK,EAAE,GAAG,CAAC,IAAI,EAAE,GAAG,CAAC,KAAK,EAAE,GAAG,CAAC,SAAS,EAAE,GAAG,CAAC,OAAO,CAAC,CAC5E,CAAC;IAEF,4EAA4E;IAC5E,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,IAAI,KAAK,CAAC,IAAI,KAAK,MAAM,IAAI,KAAK,CAAC,IAAI,KAAK,MAAM,EAAE,CAAC;YACnD,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,KAAK,EAAE,CAAC;gBAC/B,IAAI,IAAI,CAAC,IAAI,KAAK,SAAS,IAAI,IAAI,CAAC,IAAI,KAAK,MAAM,EAAE,CAAC;oBACpD,gDAAgD;oBAChD,IAAI,KAAK,CAAC,IAAI,KAAK,MAAM,EAAE,CAAC;wBAC1B,IAAI,CAAC,IAAI,GAAG,SAAS,CAAC;wBACtB,IAAI,CAAC,UAAU,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,UAAU,EAAE,GAAG,CAAC,CAAC;oBACnD,CAAC;gBACH,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAED,yEAAyE;IACzE,yDAAyD;IACzD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,IAAI,KAAK,CAAC,IAAI,KAAK,SAAS,EAAE,CAAC;YAC7B,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,KAAK,EAAE,CAAC;gBAC/B,IAAI,IAAI,CAAC,IAAI,KAAK,SAAS,EAAE,CAAC;oBAC5B,MAAM,YAAY,GAAG,YAAY,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;oBAC7C,IAAI,YAAY,CAAC,IAAI,KAAK,SAAS,EAAE,CAAC;wBACpC,IAAI,CAAC,IAAI,GAAG,YAAY,CAAC,IAAI,CAAC;wBAC9B,IAAI,CAAC,UAAU,GAAG,YAAY,CAAC,UAAU,CAAC;oBAC5C,CAAC;gBACH,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAED,MAAM,QAAQ,GAAG,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC;IAEhD,MAAM,CAAC,IAAI,CAAC,wBAAwB,MAAM,CAAC,MAAM,YAAY,QAAQ,CAAC,MAAM,QAAQ,CAAC,CAAC;IAEtF,OAAO;QACL,MAAM;QACN,OAAO,EAAE;YACP,aAAa,EAAE,MAAM,CAAC,MAAM;YAC5B,KAAK,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,MAAM,CAAC;YAC5C,KAAK,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,MAAM,CAAC;YAC5C,QAAQ,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,SAAS,CAAC;YAClD,YAAY,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,aAAa,CAAC;YAC1D,WAAW,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,YAAY,CAAC;YACxD,kBAAkB,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,qBAAqB,CAAC;YACxE,KAAK,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,MAAM,CAAC;YAC5C,OAAO,EAAE,WAAW,CAAC,MAAM,EAAE,QAAQ,EAAE,SAAS,CAAC;SAClD;KACF,CAAC;AACJ,CAAC"}

package/dist/core/parser/file-reader.d.ts

@@ -0,0 +1,13 @@
+export interface FileReadResult {
+    text: string;
+    originalName: string;
+    format: string;
+    sizeBytes: number;
+}
+/**
+ * Read a file and extract its text content based on extension.
+ * Supports: .md, .txt, .pdf, .html, .htm
+ */
+export declare function readFileContent(filePath: string, originalName?: string): Promise<FileReadResult>;
+export declare function isSupportedFormat(filename: string): boolean;
+//# sourceMappingURL=file-reader.d.ts.map

package/dist/core/parser/file-reader.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"file-reader.d.ts","sourceRoot":"","sources":["../../../src/core/parser/file-reader.ts"],"names":[],"mappings":"AAQA,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE,MAAM,CAAC;IACb,YAAY,EAAE,MAAM,CAAC;IACrB,MAAM,EAAE,MAAM,CAAC;IACf,SAAS,EAAE,MAAM,CAAC;CACnB;AAED;;;GAGG;AACH,wBAAsB,eAAe,CACnC,QAAQ,EAAE,MAAM,EAChB,YAAY,CAAC,EAAE,MAAM,GACpB,OAAO,CAAC,cAAc,CAAC,CA4CzB;AAED,wBAAgB,iBAAiB,CAAC,QAAQ,EAAE,MAAM,GAAG,OAAO,CAG3D"}

package/dist/core/parser/file-reader.js

@@ -0,0 +1,52 @@
+import { readFile } from "node:fs/promises";
+import path from "node:path";
+import { readPdfBuffer } from "./read-pdf.js";
+import { readHtmlContent } from "./read-html.js";
+import { logger } from "../utils/logger.js";
+const SUPPORTED_EXTENSIONS = new Set([".md", ".txt", ".pdf", ".html", ".htm"]);
+/**
+ * Read a file and extract its text content based on extension.
+ * Supports: .md, .txt, .pdf, .html, .htm
+ */
+export async function readFileContent(filePath, originalName) {
+    const name = originalName ?? path.basename(filePath);
+    const ext = path.extname(name).toLowerCase();
+    if (!SUPPORTED_EXTENSIONS.has(ext)) {
+        throw new Error(`Unsupported file format: "${ext}". Supported: ${[...SUPPORTED_EXTENSIONS].join(", ")}`);
+    }
+    logger.info("Reading file", { name, ext });
+    const buffer = await readFile(filePath);
+    const sizeBytes = buffer.length;
+    let text;
+    switch (ext) {
+        case ".pdf": {
+            const result = await readPdfBuffer(buffer);
+            text = result.text;
+            break;
+        }
+        case ".html":
+        case ".htm": {
+            const html = buffer.toString("utf-8");
+            text = await readHtmlContent(html);
+            break;
+        }
+        case ".md":
+        case ".txt":
+        default: {
+            text = buffer.toString("utf-8");
+            break;
+        }
+    }
+    logger.info("File read complete", { name, format: ext, textLength: text.length });
+    return {
+        text,
+        originalName: name,
+        format: ext,
+        sizeBytes,
+    };
+}
+export function isSupportedFormat(filename) {
+    const ext = path.extname(filename).toLowerCase();
+    return SUPPORTED_EXTENSIONS.has(ext);
+}
+//# sourceMappingURL=file-reader.js.map

package/dist/core/parser/file-reader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"file-reader.js","sourceRoot":"","sources":["../../../src/core/parser/file-reader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,aAAa,EAAE,MAAM,eAAe,CAAC;AAC9C,OAAO,EAAE,eAAe,EAAE,MAAM,gBAAgB,CAAC;AACjD,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAE5C,MAAM,oBAAoB,GAAG,IAAI,GAAG,CAAC,CAAC,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC;AAS/E;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,eAAe,CACnC,QAAgB,EAChB,YAAqB;IAErB,MAAM,IAAI,GAAG,YAAY,IAAI,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACrD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,WAAW,EAAE,CAAC;IAE7C,IAAI,CAAC,oBAAoB,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC;QACnC,MAAM,IAAI,KAAK,CACb,6BAA6B,GAAG,iBAAiB,CAAC,GAAG,oBAAoB,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CACxF,CAAC;IACJ,CAAC;IAED,MAAM,CAAC,IAAI,CAAC,cAAc,EAAE,EAAE,IAAI,EAAE,GAAG,EAAE,CAAC,CAAC;IAE3C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,SAAS,GAAG,MAAM,CAAC,MAAM,CAAC;IAChC,IAAI,IAAY,CAAC;IAEjB,QAAQ,GAAG,EAAE,CAAC;QACZ,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC,MAAM,CAAC,CAAC;YAC3C,IAAI,GAAG,MAAM,CAAC,IAAI,CAAC;YACnB,MAAM;QACR,CAAC;QACD,KAAK,OAAO,CAAC;QACb,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,IAAI,GAAG,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC;YACtC,IAAI,GAAG,MAAM,eAAe,CAAC,IAAI,CAAC,CAAC;YACnC,MAAM;QACR,CAAC;QACD,KAAK,KAAK,CAAC;QACX,KAAK,MAAM,CAAC;QACZ,OAAO,CAAC,CAAC,CAAC;YACR,IAAI,GAAG,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC;YAChC,MAAM;QACR,CAAC;IACH,CAAC;IAED,MAAM,CAAC,IAAI,CAAC,oBAAoB,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,GAAG,EAAE,UAAU,EAAE,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC;IAElF,OAAO;QACL,IAAI;QACJ,YAAY,EAAE,IAAI;QAClB,MAAM,EAAE,GAAG;QACX,SAAS;KACV,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,iBAAiB,CAAC,QAAgB;IAChD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IACjD,OAAO,oBAAoB,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;AACvC,CAAC"}

package/dist/core/parser/normalize.d.ts

@@ -0,0 +1,9 @@
+/**
+ * Stage 1: Text normalization.
+ * - Standardize line endings
+ * - Remove duplicate blank lines
+ * - Trim whitespace
+ * - Standardize bullet markers to "-"
+ */
+export declare function normalize(raw: string): string;
+//# sourceMappingURL=normalize.d.ts.map

package/dist/core/parser/normalize.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"normalize.d.ts","sourceRoot":"","sources":["../../../src/core/parser/normalize.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,wBAAgB,SAAS,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAiB7C"}

package/dist/core/parser/normalize.js

@@ -0,0 +1,22 @@
+/**
+ * Stage 1: Text normalization.
+ * - Standardize line endings
+ * - Remove duplicate blank lines
+ * - Trim whitespace
+ * - Standardize bullet markers to "-"
+ */
+export function normalize(raw) {
+    let text = raw.replace(/\r\n/g, "\n").replace(/\r/g, "\n");
+    // Collapse 3+ consecutive blank lines into 2
+    text = text.replace(/\n{3,}/g, "\n\n");
+    // Standardize bullet markers: *, •, ● → -
+    text = text.replace(/^(\s*)[*•●]\s/gm, "$1- ");
+    // Trim trailing whitespace per line
+    text = text
+        .split("\n")
+        .map((line) => line.trimEnd())
+        .join("\n")
+        .trim();
+    return text;
+}
+//# sourceMappingURL=normalize.js.map

package/dist/core/parser/normalize.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"normalize.js","sourceRoot":"","sources":["../../../src/core/parser/normalize.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,MAAM,UAAU,SAAS,CAAC,GAAW;IACnC,IAAI,IAAI,GAAG,GAAG,CAAC,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,IAAI,CAAC,CAAC;IAE3D,6CAA6C;IAC7C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC;IAEvC,0CAA0C;IAC1C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,iBAAiB,EAAE,MAAM,CAAC,CAAC;IAE/C,oCAAoC;IACpC,IAAI,GAAG,IAAI;SACR,KAAK,CAAC,IAAI,CAAC;SACX,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC;SAC7B,IAAI,CAAC,IAAI,CAAC;SACV,IAAI,EAAE,CAAC;IAEV,OAAO,IAAI,CAAC;AACd,CAAC"}

package/dist/core/parser/read-file.d.ts

@@ -0,0 +1,7 @@
+export interface PrdFileResult {
+    content: string;
+    absolutePath: string;
+    sizeBytes: number;
+}
+export declare function readPrdFile(filePath: string): Promise<PrdFileResult>;
+//# sourceMappingURL=read-file.d.ts.map

package/dist/core/parser/read-file.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"read-file.d.ts","sourceRoot":"","sources":["../../../src/core/parser/read-file.ts"],"names":[],"mappings":"AAKA,MAAM,WAAW,aAAa;IAC5B,OAAO,EAAE,MAAM,CAAC;IAChB,YAAY,EAAE,MAAM,CAAC;IACrB,SAAS,EAAE,MAAM,CAAC;CACnB;AAED,wBAAsB,WAAW,CAAC,QAAQ,EAAE,MAAM,GAAG,OAAO,CAAC,aAAa,CAAC,CAc1E"}

package/dist/core/parser/read-file.js

@@ -0,0 +1,17 @@
+import { readFile } from "node:fs/promises";
+import path from "node:path";
+import { fileExists } from "../utils/fs.js";
+import { FileNotFoundError } from "../utils/errors.js";
+export async function readPrdFile(filePath) {
+    const absolutePath = path.resolve(filePath);
+    if (!(await fileExists(absolutePath))) {
+        throw new FileNotFoundError(absolutePath);
+    }
+    const content = await readFile(absolutePath, "utf-8");
+    return {
+        content,
+        absolutePath,
+        sizeBytes: Buffer.byteLength(content, "utf-8"),
+    };
+}
+//# sourceMappingURL=read-file.js.map

package/dist/core/parser/read-file.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"read-file.js","sourceRoot":"","sources":["../../../src/core/parser/read-file.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,UAAU,EAAE,MAAM,gBAAgB,CAAC;AAC5C,OAAO,EAAE,iBAAiB,EAAE,MAAM,oBAAoB,CAAC;AAQvD,MAAM,CAAC,KAAK,UAAU,WAAW,CAAC,QAAgB;IAChD,MAAM,YAAY,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC;IAE5C,IAAI,CAAC,CAAC,MAAM,UAAU,CAAC,YAAY,CAAC,CAAC,EAAE,CAAC;QACtC,MAAM,IAAI,iBAAiB,CAAC,YAAY,CAAC,CAAC;IAC5C,CAAC;IAED,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,YAAY,EAAE,OAAO,CAAC,CAAC;IAEtD,OAAO;QACL,OAAO;QACP,YAAY;QACZ,SAAS,EAAE,MAAM,CAAC,UAAU,CAAC,OAAO,EAAE,OAAO,CAAC;KAC/C,CAAC;AACJ,CAAC"}

package/dist/core/parser/read-html.d.ts

@@ -0,0 +1,7 @@
+/**
+ * Extract text content from an HTML string using cheerio.
+ * Converts headings to markdown format so the parser pipeline can segment them.
+ * Strips tags, scripts, styles, and normalizes whitespace.
+ */
+export declare function readHtmlContent(html: string): Promise<string>;
+//# sourceMappingURL=read-html.d.ts.map

package/dist/core/parser/read-html.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"read-html.d.ts","sourceRoot":"","sources":["../../../src/core/parser/read-html.ts"],"names":[],"mappings":"AAWA;;;;GAIG;AACH,wBAAsB,eAAe,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CA6CnE"}