@monoes/monograph 1.1.0 → 1.2.0

package/dist/src/ingest/git-clone.js

@@ -0,0 +1,204 @@
+/**
+ * Git Clone Utility with SSRF Protection
+ *
+ * Shallow-clones repositories into ~/.monograph/repos/{name}/.
+ * If already cloned, runs `git pull --ff-only` instead.
+ *
+ * SSRF protection: only https:// and http:// are allowed,
+ * and private/internal IP ranges are blocked.
+ */
+import { spawn } from 'child_process';
+import path from 'path';
+import os from 'os';
+import fs from 'fs/promises';
+import { isIP } from 'net';
+// ── SSRF validation ───────────────────────────────────────────────────────────
+const BLOCKED_HOSTNAMES = new Set([
+    'localhost',
+    'metadata.google.internal',
+    'metadata.azure.com',
+    'metadata.internal',
+    'metadata.goog',
+]);
+/**
+ * Validate a git URL to prevent SSRF attacks.
+ *
+ * Blocks:
+ * - Non-http(s) schemes
+ * - Private IPv4 ranges (127.x, 10.x, 172.16-31.x, 192.168.x, 169.254.x, etc.)
+ * - IPv6 private ranges (::1, fc00::/7, fe80::/10, ::ffff:...)
+ * - Cloud metadata hostnames
+ * - Numeric IP encodings (decimal/hex)
+ *
+ * @throws {Error} with a descriptive message if the URL is not safe.
+ */
+export function validateGitUrl(url) {
+    let parsed;
+    try {
+        parsed = new URL(url);
+    }
+    catch {
+        throw new Error('Invalid URL');
+    }
+    if (!['https:', 'http:'].includes(parsed.protocol)) {
+        throw new Error('Only https:// and http:// git URLs are allowed');
+    }
+    const host = parsed.hostname.toLowerCase();
+    if (BLOCKED_HOSTNAMES.has(host)) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+    // Strip IPv6 brackets
+    let normalizedHost = host;
+    if (host.startsWith('[') && host.endsWith(']')) {
+        normalizedHost = host.slice(1, -1);
+    }
+    // Detect IPv6
+    const isIPv6 = isIP(normalizedHost) === 6 || normalizedHost.includes(':');
+    if (isIPv6) {
+        assertNotPrivateIPv6(normalizedHost);
+        return;
+    }
+    // Detect IPv4
+    if (isIP(normalizedHost) === 4) {
+        assertNotPrivateIPv4(normalizedHost);
+        return;
+    }
+    // Block numeric IP encodings (decimal: 2130706433, hex: 0x7f000001)
+    if (/^\d+$/.test(host) || /^0x[0-9a-f]+$/i.test(host)) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+    // Standard IPv4 dotted-notation private checks
+    if (/^127\./.test(host) ||
+        /^10\./.test(host) ||
+        /^172\.(1[6-9]|2\d|3[01])\./.test(host) ||
+        /^192\.168\./.test(host) ||
+        /^169\.254\./.test(host) ||
+        /^0\./.test(host) ||
+        host === '0.0.0.0' ||
+        /^100\.(6[4-9]|[7-9]\d|1[01]\d|12[0-7])\./.test(host) ||
+        /^198\.1[89]\./.test(host)) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+}
+function assertNotPrivateIPv6(ip) {
+    const lower = ip.toLowerCase();
+    if (lower === '::1' || lower === '0:0:0:0:0:0:0:1') {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+    if (lower === '::' || lower === '0:0:0:0:0:0:0:0') {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+    if (lower.startsWith('fc') || lower.startsWith('fd')) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+    if (lower.startsWith('fe80') ||
+        lower.startsWith('fe8') ||
+        lower.startsWith('fe9') ||
+        lower.startsWith('fea') ||
+        lower.startsWith('feb')) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+    if (lower.startsWith('::ffff:') || lower.includes(':ffff:')) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+}
+function assertNotPrivateIPv4(ip) {
+    const parts = ip.split('.').map(Number);
+    const [a, b] = parts;
+    if (a === 127 ||
+        a === 10 ||
+        (a === 172 && b >= 16 && b <= 31) ||
+        (a === 192 && b === 168) ||
+        (a === 169 && b === 254) ||
+        a === 0 ||
+        (a === 100 && b >= 64 && b <= 127) ||
+        (a === 198 && (b === 18 || b === 19))) {
+        throw new Error('Cloning from private/internal addresses is not allowed');
+    }
+}
+// ── URL helpers ────────────────────────────────────────────────────────────────
+/**
+ * Extract the repository name from an HTTPS or SSH git URL.
+ *
+ * @example
+ * extractRepoName('https://github.com/org/repo.git') // 'repo'
+ * extractRepoName('git@github.com:org/repo.git')     // 'repo'
+ */
+export function extractRepoName(url) {
+    const cleaned = url.replace(/\/+$/, '');
+    const lastSegment = cleaned.split(/[/:]/).pop() ?? 'unknown';
+    const name = lastSegment.replace(/\.git$/, '');
+    // Guard against path traversal: reject segments containing .. or path separators
+    if (!name || name.includes('..') || name.includes('/') || name.includes('\\')) {
+        return 'unknown-repo';
+    }
+    return name;
+}
+/**
+ * Get the default clone target directory for a repository name.
+ * Repositories are stored in `~/.monograph/repos/{repoName}`.
+ */
+export function getCloneDir(repoName) {
+    return path.join(os.homedir(), '.monograph', 'repos', repoName);
+}
+/**
+ * Clone or pull a git repository with SSRF protection.
+ *
+ * - If `targetDir/.git` does not exist: `git clone --depth 1 <url> <targetDir>`
+ * - If `targetDir/.git` exists: `git pull --ff-only`
+ *
+ * The URL is validated against SSRF rules before any network operation.
+ *
+ * @param url       - The git remote URL (must be https:// or http://).
+ * @param targetDir - Local directory to clone into.
+ * @param onProgress - Optional progress callback.
+ * @returns Resolves to `targetDir` on success.
+ */
+export async function cloneOrPull(url, targetDir, onProgress) {
+    // Always validate before any operation
+    validateGitUrl(url);
+    const hasGitDir = await fs.access(path.join(targetDir, '.git')).then(() => true, () => false);
+    if (hasGitDir) {
+        onProgress?.({ phase: 'pulling', message: 'Pulling latest changes…' });
+        await runGit(['pull', '--ff-only'], targetDir);
+    }
+    else {
+        await fs.mkdir(path.dirname(targetDir), { recursive: true });
+        onProgress?.({ phase: 'cloning', message: `Cloning ${url}…` });
+        await runGit(['clone', '--depth', '1', url, targetDir]);
+    }
+    return targetDir;
+}
+// ── git subprocess ─────────────────────────────────────────────────────────────
+function runGit(args, cwd) {
+    return new Promise((resolve, reject) => {
+        const proc = spawn('git', args, {
+            cwd,
+            stdio: ['ignore', 'pipe', 'pipe'],
+            env: {
+                ...process.env,
+                GIT_TERMINAL_PROMPT: '0',
+                GIT_ASKPASS: process.platform === 'win32' ? 'echo' : '/bin/true',
+            },
+        });
+        let stderr = '';
+        proc.stderr.on('data', (chunk) => {
+            stderr += chunk;
+        });
+        proc.on('close', code => {
+            if (code === 0) {
+                resolve();
+            }
+            else {
+                if (stderr.trim()) {
+                    console.error(`git ${args[0]} stderr: ${stderr.trim()}`);
+                }
+                reject(new Error(`git ${args[0]} failed (exit code ${code})`));
+            }
+        });
+        proc.on('error', err => {
+            reject(new Error(`Failed to spawn git: ${err.message}`));
+        });
+    });
+}
+//# sourceMappingURL=git-clone.js.map

package/dist/src/ingest/git-clone.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"git-clone.js","sourceRoot":"","sources":["../../../src/ingest/git-clone.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,EAAE,KAAK,EAAE,MAAM,eAAe,CAAC;AACtC,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,EAAE,MAAM,aAAa,CAAC;AAC7B,OAAO,EAAE,IAAI,EAAE,MAAM,KAAK,CAAC;AAE3B,iFAAiF;AAEjF,MAAM,iBAAiB,GAAG,IAAI,GAAG,CAAC;IAChC,WAAW;IACX,0BAA0B;IAC1B,oBAAoB;IACpB,mBAAmB;IACnB,eAAe;CAChB,CAAC,CAAC;AAEH;;;;;;;;;;;GAWG;AACH,MAAM,UAAU,cAAc,CAAC,GAAW;IACxC,IAAI,MAAW,CAAC;IAChB,IAAI,CAAC;QACH,MAAM,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACxB,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,KAAK,CAAC,aAAa,CAAC,CAAC;IACjC,CAAC;IAED,IAAI,CAAC,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,QAAQ,CAAC,EAAE,CAAC;QACnD,MAAM,IAAI,KAAK,CAAC,gDAAgD,CAAC,CAAC;IACpE,CAAC;IAED,MAAM,IAAI,GAAG,MAAM,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC;IAE3C,IAAI,iBAAiB,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC;QAChC,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;IAED,sBAAsB;IACtB,IAAI,cAAc,GAAG,IAAI,CAAC;IAC1B,IAAI,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,IAAI,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;QAC/C,cAAc,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IACrC,CAAC;IAED,cAAc;IACd,MAAM,MAAM,GAAG,IAAI,CAAC,cAAc,CAAC,KAAK,CAAC,IAAI,cAAc,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;IAC1E,IAAI,MAAM,EAAE,CAAC;QACX,oBAAoB,CAAC,cAAc,CAAC,CAAC;QACrC,OAAO;IACT,CAAC;IAED,cAAc;IACd,IAAI,IAAI,CAAC,cAAc,CAAC,KAAK,CAAC,EAAE,CAAC;QAC/B,oBAAoB,CAAC,cAAc,CAAC,CAAC;QACrC,OAAO;IACT,CAAC;IAED,oEAAoE;IACpE,IAAI,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,gBAAgB,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;QACtD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;IAED,+CAA+C;IAC/C,IACE,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC;QACnB,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAClB,4BAA4B,CAAC,IAAI,CAAC,IAAI,CAAC;QACvC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC;QACxB,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC;QACxB,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC;QACjB,IAAI,KAAK,SAAS;QAClB,0CAA0C,CAAC,IAAI,CAAC,IAAI,CAAC;QACrD,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,EAC1B,CAAC;QACD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;AACH,CAAC;AAED,SAAS,oBAAoB,CAAC,EAAU;IACtC,MAAM,KAAK,GAAG,EAAE,CAAC,WAAW,EAAE,CAAC;IAC/B,IAAI,KAAK,KAAK,KAAK,IAAI,KAAK,KAAK,iBAAiB,EAAE,CAAC;QACnD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;IACD,IAAI,KAAK,KAAK,IAAI,IAAI,KAAK,KAAK,iBAAiB,EAAE,CAAC;QAClD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;IACD,IAAI,KAAK,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,KAAK,CAAC,UAAU,CAAC,IAAI,CAAC,EAAE,CAAC;QACrD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;IACD,IACE,KAAK,CAAC,UAAU,CAAC,MAAM,CAAC;QACxB,KAAK,CAAC,UAAU,CAAC,KAAK,CAAC;QACvB,KAAK,CAAC,UAAU,CAAC,KAAK,CAAC;QACvB,KAAK,CAAC,UAAU,CAAC,KAAK,CAAC;QACvB,KAAK,CAAC,UAAU,CAAC,KAAK,CAAC,EACvB,CAAC;QACD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;IACD,IAAI,KAAK,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC5D,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;AACH,CAAC;AAED,SAAS,oBAAoB,CAAC,EAAU;IACtC,MAAM,KAAK,GAAG,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;IACxC,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,GAAG,KAAK,CAAC;IACrB,IACE,CAAC,KAAK,GAAG;QACT,CAAC,KAAK,EAAE;QACR,CAAC,CAAC,KAAK,GAAG,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,IAAI,EAAE,CAAC;QACjC,CAAC,CAAC,KAAK,GAAG,IAAI,CAAC,KAAK,GAAG,CAAC;QACxB,CAAC,CAAC,KAAK,GAAG,IAAI,CAAC,KAAK,GAAG,CAAC;QACxB,CAAC,KAAK,CAAC;QACP,CAAC,CAAC,KAAK,GAAG,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,IAAI,GAAG,CAAC;QAClC,CAAC,CAAC,KAAK,GAAG,IAAI,CAAC,CAAC,KAAK,EAAE,IAAI,CAAC,KAAK,EAAE,CAAC,CAAC,EACrC,CAAC;QACD,MAAM,IAAI,KAAK,CAAC,wDAAwD,CAAC,CAAC;IAC5E,CAAC;AACH,CAAC;AAED,kFAAkF;AAElF;;;;;;GAMG;AACH,MAAM,UAAU,eAAe,CAAC,GAAW;IACzC,MAAM,OAAO,GAAG,GAAG,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;IACxC,MAAM,WAAW,GAAG,OAAO,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,IAAI,SAAS,CAAC;IAC7D,MAAM,IAAI,GAAG,WAAW,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC;IAC/C,iFAAiF;IACjF,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC;QAC9E,OAAO,cAAc,CAAC;IACxB,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,WAAW,CAAC,QAAgB;IAC1C,OAAO,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,EAAE,YAAY,EAAE,OAAO,EAAE,QAAQ,CAAC,CAAC;AAClE,CAAC;AASD;;;;;;;;;;;;GAYG;AACH,MAAM,CAAC,KAAK,UAAU,WAAW,CAC/B,GAAW,EACX,SAAiB,EACjB,UAA8C;IAE9C,uCAAuC;IACvC,cAAc,CAAC,GAAG,CAAC,CAAC;IAEpB,MAAM,SAAS,GAAG,MAAM,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC,CAAC,IAAI,CAClE,GAAG,EAAE,CAAC,IAAI,EACV,GAAG,EAAE,CAAC,KAAK,CACZ,CAAC;IAEF,IAAI,SAAS,EAAE,CAAC;QACd,UAAU,EAAE,CAAC,EAAE,KAAK,EAAE,SAAS,EAAE,OAAO,EAAE,yBAAyB,EAAE,CAAC,CAAC;QACvE,MAAM,MAAM,CAAC,CAAC,MAAM,EAAE,WAAW,CAAC,EAAE,SAAS,CAAC,CAAC;IACjD,CAAC;SAAM,CAAC;QACN,MAAM,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,SAAS,CAAC,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QAC7D,UAAU,EAAE,CAAC,EAAE,KAAK,EAAE,SAAS,EAAE,OAAO,EAAE,WAAW,GAAG,GAAG,EAAE,CAAC,CAAC;QAC/D,MAAM,MAAM,CAAC,CAAC,OAAO,EAAE,SAAS,EAAE,GAAG,EAAE,GAAG,EAAE,SAAS,CAAC,CAAC,CAAC;IAC1D,CAAC;IAED,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,kFAAkF;AAElF,SAAS,MAAM,CAAC,IAAc,EAAE,GAAY;IAC1C,OAAO,IAAI,OAAO,CAAC,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;QACrC,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,EAAE,IAAI,EAAE;YAC9B,GAAG;YACH,KAAK,EAAE,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,CAAC;YACjC,GAAG,EAAE;gBACH,GAAG,OAAO,CAAC,GAAG;gBACd,mBAAmB,EAAE,GAAG;gBACxB,WAAW,EAAE,OAAO,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,WAAW;aACjE;SACF,CAAC,CAAC;QAEH,IAAI,MAAM,GAAG,EAAE,CAAC;QAChB,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,CAAC,KAAa,EAAE,EAAE;YACvC,MAAM,IAAI,KAAK,CAAC;QAClB,CAAC,CAAC,CAAC;QAEH,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,IAAI,CAAC,EAAE;YACtB,IAAI,IAAI,KAAK,CAAC,EAAE,CAAC;gBACf,OAAO,EAAE,CAAC;YACZ,CAAC;iBAAM,CAAC;gBACN,IAAI,MAAM,CAAC,IAAI,EAAE,EAAE,CAAC;oBAClB,OAAO,CAAC,KAAK,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,YAAY,MAAM,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC;gBAC3D,CAAC;gBACD,MAAM,CAAC,IAAI,KAAK,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,sBAAsB,IAAI,GAAG,CAAC,CAAC,CAAC;YACjE,CAAC;QACH,CAAC,CAAC,CAAC;QAEH,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,GAAG,CAAC,EAAE;YACrB,MAAM,CAAC,IAAI,KAAK,CAAC,wBAAwB,GAAG,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QAC3D,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;AACL,CAAC"}

package/dist/src/ingest/query-memory.d.ts

@@ -0,0 +1,48 @@
+/**
+ * Q&A memory feedback loop.
+ *
+ * Mirrors graphify's `save_query_result`: saves question/answer pairs as
+ * Markdown files with YAML front-matter into a memory directory.  On the next
+ * rebuild those files are treated as document nodes, so the graph grows smarter
+ * from both what you add AND what you ask.
+ *
+ * Intended for use after `monographQuery` / `monographExplain` calls so that
+ * insights are persisted and available during future pipeline runs.
+ */
+export interface SaveQueryResultOptions {
+    /** Question / query string. */
+    question: string;
+    /** Answer produced by the graph query. */
+    answer: string;
+    /**
+     * Directory where memory files are written.
+     * Defaults to `<cwd>/monograph-out/memory`.
+     */
+    memoryDir?: string;
+    /** Free-form query type label (e.g. "query", "explain", "impact"). Default: "query". */
+    queryType?: string;
+    /**
+     * Up to 10 node ids / names that were the primary sources for the answer.
+     * These are stored in front-matter so a future extraction pass can wire up
+     * edges to the originating nodes.
+     */
+    sourceNodes?: string[];
+}
+export interface SaveQueryResultOutput {
+    /** Absolute path of the written Markdown file. */
+    filePath: string;
+    /** ISO-8601 timestamp of when the file was written. */
+    writtenAt: string;
+}
+/**
+ * Save a Q&A query result as a Markdown file in the memory directory.
+ *
+ * The file has YAML front-matter that monograph's document extractor reads
+ * as node metadata.  The body contains the question and answer in a
+ * human-readable format.
+ *
+ * @param options - Query options including question, answer, and target dir.
+ * @returns The file path and write timestamp.
+ */
+export declare function saveQueryResult(options: SaveQueryResultOptions): SaveQueryResultOutput;
+//# sourceMappingURL=query-memory.d.ts.map

package/dist/src/ingest/query-memory.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"query-memory.d.ts","sourceRoot":"","sources":["../../../src/ingest/query-memory.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAOH,MAAM,WAAW,sBAAsB;IACrC,+BAA+B;IAC/B,QAAQ,EAAE,MAAM,CAAC;IACjB,0CAA0C;IAC1C,MAAM,EAAE,MAAM,CAAC;IACf;;;OAGG;IACH,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,wFAAwF;IACxF,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB;;;;OAIG;IACH,WAAW,CAAC,EAAE,MAAM,EAAE,CAAC;CACxB;AAED,MAAM,WAAW,qBAAqB;IACpC,kDAAkD;IAClD,QAAQ,EAAE,MAAM,CAAC;IACjB,uDAAuD;IACvD,SAAS,EAAE,MAAM,CAAC;CACnB;AA0BD;;;;;;;;;GASG;AACH,wBAAgB,eAAe,CAAC,OAAO,EAAE,sBAAsB,GAAG,qBAAqB,CAqDtF"}

package/dist/src/ingest/query-memory.js

@@ -0,0 +1,81 @@
+/**
+ * Q&A memory feedback loop.
+ *
+ * Mirrors graphify's `save_query_result`: saves question/answer pairs as
+ * Markdown files with YAML front-matter into a memory directory.  On the next
+ * rebuild those files are treated as document nodes, so the graph grows smarter
+ * from both what you add AND what you ask.
+ *
+ * Intended for use after `monographQuery` / `monographExplain` calls so that
+ * insights are persisted and available during future pipeline runs.
+ */
+import { mkdirSync, writeFileSync } from 'fs';
+import { join } from 'path';
+// ── Helpers ────────────────────────────────────────────────────────────────────
+function yamlStr(s) {
+    return s.replace(/\\/g, '\\\\').replace(/"/g, '\\"').replace(/\n/g, ' ').replace(/\r/g, '');
+}
+function slug(question) {
+    return question
+        .toLowerCase()
+        .replace(/[^\w]/g, '_')
+        .slice(0, 50)
+        .replace(/^_+|_+$/g, '');
+}
+function isoNow() {
+    return new Date().toISOString();
+}
+function timestamp() {
+    return new Date().toISOString().replace(/[-:T]/g, '').replace(/\..+/, '').slice(0, 15);
+}
+// ── Main export ────────────────────────────────────────────────────────────────
+/**
+ * Save a Q&A query result as a Markdown file in the memory directory.
+ *
+ * The file has YAML front-matter that monograph's document extractor reads
+ * as node metadata.  The body contains the question and answer in a
+ * human-readable format.
+ *
+ * @param options - Query options including question, answer, and target dir.
+ * @returns The file path and write timestamp.
+ */
+export function saveQueryResult(options) {
+    const { question, answer, memoryDir = join(process.cwd(), 'monograph-out', 'memory'), queryType = 'query', sourceNodes, } = options;
+    mkdirSync(memoryDir, { recursive: true });
+    const now = isoNow();
+    const ts = timestamp();
+    const s = slug(question);
+    const filename = `query_${ts}_${s}.md`;
+    const filePath = join(memoryDir, filename);
+    const frontmatterLines = [
+        '---',
+        `type: "${queryType}"`,
+        `date: "${now}"`,
+        `question: "${yamlStr(question)}"`,
+        'contributor: "monograph"',
+    ];
+    if (sourceNodes && sourceNodes.length > 0) {
+        const limited = sourceNodes.slice(0, 10);
+        const nodesStr = limited.map(n => `"${yamlStr(n)}"`).join(', ');
+        frontmatterLines.push(`source_nodes: [${nodesStr}]`);
+    }
+    frontmatterLines.push('---');
+    const bodyLines = [
+        '',
+        `# Q: ${question}`,
+        '',
+        '## Answer',
+        '',
+        answer,
+    ];
+    if (sourceNodes && sourceNodes.length > 0) {
+        bodyLines.push('', '## Source Nodes', '');
+        for (const n of sourceNodes) {
+            bodyLines.push(`- ${n}`);
+        }
+    }
+    const content = [...frontmatterLines, ...bodyLines].join('\n');
+    writeFileSync(filePath, content, 'utf-8');
+    return { filePath, writtenAt: now };
+}
+//# sourceMappingURL=query-memory.js.map

package/dist/src/ingest/query-memory.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"query-memory.js","sourceRoot":"","sources":["../../../src/ingest/query-memory.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAEH,OAAO,EAAE,SAAS,EAAE,aAAa,EAAE,MAAM,IAAI,CAAC;AAC9C,OAAO,EAAE,IAAI,EAAE,MAAM,MAAM,CAAC;AA+B5B,kFAAkF;AAElF,SAAS,OAAO,CAAC,CAAS;IACxB,OAAO,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC,OAAO,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;AAC9F,CAAC;AAED,SAAS,IAAI,CAAC,QAAgB;IAC5B,OAAO,QAAQ;SACZ,WAAW,EAAE;SACb,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC;SACtB,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC;SACZ,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC;AAC7B,CAAC;AAED,SAAS,MAAM;IACb,OAAO,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;AAClC,CAAC;AAED,SAAS,SAAS;IAChB,OAAO,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;AACzF,CAAC;AAED,kFAAkF;AAElF;;;;;;;;;GASG;AACH,MAAM,UAAU,eAAe,CAAC,OAA+B;IAC7D,MAAM,EACJ,QAAQ,EACR,MAAM,EACN,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,GAAG,EAAE,EAAE,eAAe,EAAE,QAAQ,CAAC,EAC1D,SAAS,GAAG,OAAO,EACnB,WAAW,GACZ,GAAG,OAAO,CAAC;IAEZ,SAAS,CAAC,SAAS,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAE1C,MAAM,GAAG,GAAG,MAAM,EAAE,CAAC;IACrB,MAAM,EAAE,GAAG,SAAS,EAAE,CAAC;IACvB,MAAM,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC;IACzB,MAAM,QAAQ,GAAG,SAAS,EAAE,IAAI,CAAC,KAAK,CAAC;IACvC,MAAM,QAAQ,GAAG,IAAI,CAAC,SAAS,EAAE,QAAQ,CAAC,CAAC;IAE3C,MAAM,gBAAgB,GAAG;QACvB,KAAK;QACL,UAAU,SAAS,GAAG;QACtB,UAAU,GAAG,GAAG;QAChB,cAAc,OAAO,CAAC,QAAQ,CAAC,GAAG;QAClC,0BAA0B;KAC3B,CAAC;IAEF,IAAI,WAAW,IAAI,WAAW,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC1C,MAAM,OAAO,GAAG,WAAW,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QACzC,MAAM,QAAQ,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,IAAI,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAChE,gBAAgB,CAAC,IAAI,CAAC,kBAAkB,QAAQ,GAAG,CAAC,CAAC;IACvD,CAAC;IAED,gBAAgB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAE7B,MAAM,SAAS,GAAG;QAChB,EAAE;QACF,QAAQ,QAAQ,EAAE;QAClB,EAAE;QACF,WAAW;QACX,EAAE;QACF,MAAM;KACP,CAAC;IAEF,IAAI,WAAW,IAAI,WAAW,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC1C,SAAS,CAAC,IAAI,CAAC,EAAE,EAAE,iBAAiB,EAAE,EAAE,CAAC,CAAC;QAC1C,KAAK,MAAM,CAAC,IAAI,WAAW,EAAE,CAAC;YAC5B,SAAS,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;QAC3B,CAAC;IACH,CAAC;IAED,MAAM,OAAO,GAAG,CAAC,GAAG,gBAAgB,EAAE,GAAG,SAAS,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAC/D,aAAa,CAAC,QAAQ,EAAE,OAAO,EAAE,OAAO,CAAC,CAAC;IAE1C,OAAO,EAAE,QAAQ,EAAE,SAAS,EAAE,GAAG,EAAE,CAAC;AACtC,CAAC"}

package/dist/src/ingest/transcribe.d.ts

@@ -0,0 +1,71 @@
+/**
+ * Video / audio transcription support.
+ *
+ * Mirrors graphify's `transcribe.py`: converts video/audio files or URLs to
+ * plain-text transcript files that the pipeline can then extract as document
+ * nodes.
+ *
+ * Implementation strategy
+ * -----------------------
+ * TypeScript cannot import Python's faster-whisper directly.  Instead this
+ * module shells out to the faster-whisper CLI (`faster-whisper`) or the
+ * `whisper` CLI (OpenAI's original).  yt-dlp is used for URL audio downloads.
+ * If neither tool is installed, functions throw descriptive errors.
+ *
+ * Environment variables
+ * ---------------------
+ * MONOGRAPH_WHISPER_MODEL   - Whisper model name (default: "base")
+ * MONOGRAPH_WHISPER_PROMPT  - Override initial prompt for Whisper
+ */
+export declare const VIDEO_EXTENSIONS: Set<string>;
+export declare function isUrl(input: string): boolean;
+/**
+ * Download audio from a URL using yt-dlp.
+ * Returns the path to the downloaded audio file.
+ * Cached: if a file with the same URL hash already exists it is returned immediately.
+ *
+ * @throws {Error} if yt-dlp is not installed.
+ */
+export declare function downloadAudio(url: string, outputDir: string): Promise<string>;
+/**
+ * Build a domain hint for Whisper from a set of god-node labels.
+ * Mirrors graphify's `build_whisper_prompt`.
+ */
+export declare function buildWhisperPrompt(godNodeLabels: string[]): string;
+export interface TranscribeOptions {
+    /** Directory to write transcript files. Default: monograph-out/transcripts */
+    outputDir?: string;
+    /** Initial prompt / domain hint passed to Whisper. */
+    initialPrompt?: string;
+    /** Re-transcribe even if transcript already exists. Default: false */
+    force?: boolean;
+}
+export interface TranscribeResult {
+    transcriptPath: string;
+    /** Number of non-empty lines in the transcript. */
+    lineCount: number;
+    /** Whether the transcript was loaded from cache (not freshly generated). */
+    fromCache: boolean;
+}
+/**
+ * Transcribe a video/audio file or URL to a plain-text transcript.
+ *
+ * Requires either `faster-whisper` or `whisper` CLI to be on the PATH.
+ * For URLs, also requires `yt-dlp`.
+ *
+ * @param input   - A local file path or a URL (http/https/www).
+ * @param options - Transcription options.
+ * @returns Resolved transcript path and metadata.
+ */
+export declare function transcribe(input: string, options?: TranscribeOptions): Promise<TranscribeResult>;
+/**
+ * Transcribe a list of video/audio files or URLs.
+ * Already-transcribed files are returned from cache instantly.
+ * The `initialPrompt` is shared across all files.
+ *
+ * @param inputs  - Array of file paths or URLs.
+ * @param options - Shared transcription options.
+ * @returns Array of TranscribeResult (one per input, in order).
+ */
+export declare function transcribeAll(inputs: string[], options?: TranscribeOptions): Promise<TranscribeResult[]>;
+//# sourceMappingURL=transcribe.d.ts.map

package/dist/src/ingest/transcribe.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"transcribe.d.ts","sourceRoot":"","sources":["../../../src/ingest/transcribe.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAaH,eAAO,MAAM,gBAAgB,aAG3B,CAAC;AASH,wBAAgB,KAAK,CAAC,KAAK,EAAE,MAAM,GAAG,OAAO,CAE5C;AA6BD;;;;;;GAMG;AACH,wBAAsB,aAAa,CAAC,GAAG,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAqCnF;AAID;;;GAGG;AACH,wBAAgB,kBAAkB,CAAC,aAAa,EAAE,MAAM,EAAE,GAAG,MAAM,CAQlE;AAID,MAAM,WAAW,iBAAiB;IAChC,8EAA8E;IAC9E,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,sDAAsD;IACtD,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,sEAAsE;IACtE,KAAK,CAAC,EAAE,OAAO,CAAC;CACjB;AAED,MAAM,WAAW,gBAAgB;IAC/B,cAAc,EAAE,MAAM,CAAC;IACvB,mDAAmD;IACnD,SAAS,EAAE,MAAM,CAAC;IAClB,4EAA4E;IAC5E,SAAS,EAAE,OAAO,CAAC;CACpB;AAED;;;;;;;;;GASG;AACH,wBAAsB,UAAU,CAC9B,KAAK,EAAE,MAAM,EACb,OAAO,GAAE,iBAAsB,GAC9B,OAAO,CAAC,gBAAgB,CAAC,CAyD3B;AAID;;;;;;;;GAQG;AACH,wBAAsB,aAAa,CACjC,MAAM,EAAE,MAAM,EAAE,EAChB,OAAO,GAAE,iBAAsB,GAC9B,OAAO,CAAC,gBAAgB,EAAE,CAAC,CAY7B"}

package/dist/src/ingest/transcribe.js

@@ -0,0 +1,199 @@
+/**
+ * Video / audio transcription support.
+ *
+ * Mirrors graphify's `transcribe.py`: converts video/audio files or URLs to
+ * plain-text transcript files that the pipeline can then extract as document
+ * nodes.
+ *
+ * Implementation strategy
+ * -----------------------
+ * TypeScript cannot import Python's faster-whisper directly.  Instead this
+ * module shells out to the faster-whisper CLI (`faster-whisper`) or the
+ * `whisper` CLI (OpenAI's original).  yt-dlp is used for URL audio downloads.
+ * If neither tool is installed, functions throw descriptive errors.
+ *
+ * Environment variables
+ * ---------------------
+ * MONOGRAPH_WHISPER_MODEL   - Whisper model name (default: "base")
+ * MONOGRAPH_WHISPER_PROMPT  - Override initial prompt for Whisper
+ */
+import { execFile } from 'child_process';
+import { promisify } from 'util';
+import { createHash } from 'crypto';
+import { mkdirSync, existsSync, writeFileSync } from 'fs';
+import { readFile } from 'fs/promises';
+import path from 'path';
+const execFileAsync = promisify(execFile);
+// ── Constants ──────────────────────────────────────────────────────────────────
+export const VIDEO_EXTENSIONS = new Set([
+    '.mp4', '.mov', '.webm', '.mkv', '.avi', '.m4v',
+    '.mp3', '.wav', '.m4a', '.ogg',
+]);
+const URL_PREFIXES = ['http://', 'https://', 'www.'];
+const DEFAULT_MODEL = 'base';
+const FALLBACK_PROMPT = 'Use proper punctuation and paragraph breaks.';
+const DEFAULT_TRANSCRIPTS_DIR = 'monograph-out/transcripts';
+// ── Helpers ────────────────────────────────────────────────────────────────────
+export function isUrl(input) {
+    return URL_PREFIXES.some(p => input.startsWith(p));
+}
+function getModel() {
+    return process.env['MONOGRAPH_WHISPER_MODEL'] ?? DEFAULT_MODEL;
+}
+async function whisperCliAvailable() {
+    for (const cmd of ['faster-whisper', 'whisper']) {
+        try {
+            await execFileAsync(cmd, ['--help']);
+            return cmd;
+        }
+        catch {
+            // not found — try next
+        }
+    }
+    return null;
+}
+async function ytdlpAvailable() {
+    try {
+        await execFileAsync('yt-dlp', ['--version']);
+        return true;
+    }
+    catch {
+        return false;
+    }
+}
+// ── Download audio ─────────────────────────────────────────────────────────────
+/**
+ * Download audio from a URL using yt-dlp.
+ * Returns the path to the downloaded audio file.
+ * Cached: if a file with the same URL hash already exists it is returned immediately.
+ *
+ * @throws {Error} if yt-dlp is not installed.
+ */
+export async function downloadAudio(url, outputDir) {
+    if (!await ytdlpAvailable()) {
+        throw new Error('Video URL download requires yt-dlp. Install it with: pip install yt-dlp');
+    }
+    mkdirSync(outputDir, { recursive: true });
+    const urlHash = createHash('sha1').update(url).digest('hex').slice(0, 12);
+    const audioExts = ['.m4a', '.opus', '.mp3', '.ogg', '.wav', '.webm'];
+    // Return cached file if present
+    for (const ext of audioExts) {
+        const candidate = path.join(outputDir, `yt_${urlHash}${ext}`);
+        if (existsSync(candidate)) {
+            return candidate;
+        }
+    }
+    const outTemplate = path.join(outputDir, `yt_${urlHash}.%(ext)s`);
+    await execFileAsync('yt-dlp', [
+        '--format', 'bestaudio[ext=m4a]/bestaudio/best',
+        '--output', outTemplate,
+        '--quiet',
+        '--no-warnings',
+        '--no-playlist',
+        url,
+    ]);
+    // Find the downloaded file
+    for (const ext of audioExts) {
+        const candidate = path.join(outputDir, `yt_${urlHash}${ext}`);
+        if (existsSync(candidate))
+            return candidate;
+    }
+    throw new Error(`yt-dlp download completed but output file not found for URL: ${url}`);
+}
+// ── Whisper prompt ─────────────────────────────────────────────────────────────
+/**
+ * Build a domain hint for Whisper from a set of god-node labels.
+ * Mirrors graphify's `build_whisper_prompt`.
+ */
+export function buildWhisperPrompt(godNodeLabels) {
+    const override = process.env['MONOGRAPH_WHISPER_PROMPT'];
+    if (override)
+        return override;
+    if (godNodeLabels.length === 0)
+        return FALLBACK_PROMPT;
+    const topics = godNodeLabels.slice(0, 5).join(', ');
+    return `Technical discussion about ${topics}. Use proper punctuation and paragraph breaks.`;
+}
+/**
+ * Transcribe a video/audio file or URL to a plain-text transcript.
+ *
+ * Requires either `faster-whisper` or `whisper` CLI to be on the PATH.
+ * For URLs, also requires `yt-dlp`.
+ *
+ * @param input   - A local file path or a URL (http/https/www).
+ * @param options - Transcription options.
+ * @returns Resolved transcript path and metadata.
+ */
+export async function transcribe(input, options = {}) {
+    const outDir = options.outputDir ?? DEFAULT_TRANSCRIPTS_DIR;
+    mkdirSync(outDir, { recursive: true });
+    // Resolve audio path
+    let audioPath;
+    if (isUrl(input)) {
+        const dlDir = path.join(outDir, 'downloads');
+        audioPath = await downloadAudio(input, dlDir);
+    }
+    else {
+        audioPath = input;
+    }
+    const stem = path.basename(audioPath, path.extname(audioPath));
+    const transcriptPath = path.join(outDir, `${stem}.txt`);
+    // Return cached transcript if it exists
+    if (!options.force && existsSync(transcriptPath)) {
+        const content = await readFile(transcriptPath, 'utf-8');
+        const lines = content.split('\n').filter(l => l.trim().length > 0);
+        return { transcriptPath, lineCount: lines.length, fromCache: true };
+    }
+    // Find available Whisper CLI
+    const whisperCmd = await whisperCliAvailable();
+    if (!whisperCmd) {
+        throw new Error('Video transcription requires faster-whisper or whisper CLI. ' +
+            'Install with: pip install faster-whisper  or  pip install openai-whisper');
+    }
+    const model = getModel();
+    const prompt = options.initialPrompt ?? FALLBACK_PROMPT;
+    const args = [audioPath, '--model', model, '--output_format', 'txt', '--output_dir', outDir];
+    if (prompt)
+        args.push('--initial_prompt', prompt);
+    await execFileAsync(whisperCmd, args);
+    // Whisper may write <stem>.txt directly or <stem>.<ext>.txt — normalise
+    if (!existsSync(transcriptPath)) {
+        // Try with original extension included in stem
+        const altPath = path.join(outDir, `${path.basename(audioPath)}.txt`);
+        if (existsSync(altPath)) {
+            // Rename to canonical path
+            const content = await readFile(altPath, 'utf-8');
+            writeFileSync(transcriptPath, content, 'utf-8');
+        }
+    }
+    const content = existsSync(transcriptPath)
+        ? await readFile(transcriptPath, 'utf-8')
+        : '';
+    const lines = content.split('\n').filter(l => l.trim().length > 0);
+    return { transcriptPath, lineCount: lines.length, fromCache: false };
+}
+// ── Transcribe many ────────────────────────────────────────────────────────────
+/**
+ * Transcribe a list of video/audio files or URLs.
+ * Already-transcribed files are returned from cache instantly.
+ * The `initialPrompt` is shared across all files.
+ *
+ * @param inputs  - Array of file paths or URLs.
+ * @param options - Shared transcription options.
+ * @returns Array of TranscribeResult (one per input, in order).
+ */
+export async function transcribeAll(inputs, options = {}) {
+    if (inputs.length === 0)
+        return [];
+    const results = [];
+    for (const input of inputs) {
+        try {
+            results.push(await transcribe(input, options));
+        }
+        catch (err) {
+            console.error(`[monograph transcribe] Failed for ${input}: ${err.message}`);
+        }
+    }
+    return results;
+}
+//# sourceMappingURL=transcribe.js.map

package/dist/src/ingest/transcribe.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"transcribe.js","sourceRoot":"","sources":["../../../src/ingest/transcribe.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAEH,OAAO,EAAE,QAAQ,EAAE,MAAM,eAAe,CAAC;AACzC,OAAO,EAAE,SAAS,EAAE,MAAM,MAAM,CAAC;AACjC,OAAO,EAAE,UAAU,EAAE,MAAM,QAAQ,CAAC;AACpC,OAAO,EAAE,SAAS,EAAE,UAAU,EAAE,aAAa,EAAE,MAAM,IAAI,CAAC;AAC1D,OAAO,EAAE,QAAQ,EAAa,MAAM,aAAa,CAAC;AAClD,OAAO,IAAI,MAAM,MAAM,CAAC;AAExB,MAAM,aAAa,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;AAE1C,kFAAkF;AAElF,MAAM,CAAC,MAAM,gBAAgB,GAAG,IAAI,GAAG,CAAC;IACtC,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM;IAC/C,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM;CAC/B,CAAC,CAAC;AAEH,MAAM,YAAY,GAAG,CAAC,SAAS,EAAE,UAAU,EAAE,MAAM,CAAC,CAAC;AACrD,MAAM,aAAa,GAAG,MAAM,CAAC;AAC7B,MAAM,eAAe,GAAG,8CAA8C,CAAC;AACvE,MAAM,uBAAuB,GAAG,2BAA2B,CAAC;AAE5D,kFAAkF;AAElF,MAAM,UAAU,KAAK,CAAC,KAAa;IACjC,OAAO,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC;AACrD,CAAC;AAED,SAAS,QAAQ;IACf,OAAO,OAAO,CAAC,GAAG,CAAC,yBAAyB,CAAC,IAAI,aAAa,CAAC;AACjE,CAAC;AAED,KAAK,UAAU,mBAAmB;IAChC,KAAK,MAAM,GAAG,IAAI,CAAC,gBAAgB,EAAE,SAAS,CAAC,EAAE,CAAC;QAChD,IAAI,CAAC;YACH,MAAM,aAAa,CAAC,GAAG,EAAE,CAAC,QAAQ,CAAC,CAAC,CAAC;YACrC,OAAO,GAAG,CAAC;QACb,CAAC;QAAC,MAAM,CAAC;YACP,uBAAuB;QACzB,CAAC;IACH,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,KAAK,UAAU,cAAc;IAC3B,IAAI,CAAC;QACH,MAAM,aAAa,CAAC,QAAQ,EAAE,CAAC,WAAW,CAAC,CAAC,CAAC;QAC7C,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED,kFAAkF;AAElF;;;;;;GAMG;AACH,MAAM,CAAC,KAAK,UAAU,aAAa,CAAC,GAAW,EAAE,SAAiB;IAChE,IAAI,CAAC,MAAM,cAAc,EAAE,EAAE,CAAC;QAC5B,MAAM,IAAI,KAAK,CACb,yEAAyE,CAC1E,CAAC;IACJ,CAAC;IAED,SAAS,CAAC,SAAS,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAE1C,MAAM,OAAO,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IAC1E,MAAM,SAAS,GAAG,CAAC,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC;IAErE,gCAAgC;IAChC,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC5B,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,MAAM,OAAO,GAAG,GAAG,EAAE,CAAC,CAAC;QAC9D,IAAI,UAAU,CAAC,SAAS,CAAC,EAAE,CAAC;YAC1B,OAAO,SAAS,CAAC;QACnB,CAAC;IACH,CAAC;IAED,MAAM,WAAW,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,MAAM,OAAO,UAAU,CAAC,CAAC;IAClE,MAAM,aAAa,CAAC,QAAQ,EAAE;QAC5B,UAAU,EAAE,mCAAmC;QAC/C,UAAU,EAAE,WAAW;QACvB,SAAS;QACT,eAAe;QACf,eAAe;QACf,GAAG;KACJ,CAAC,CAAC;IAEH,2BAA2B;IAC3B,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC5B,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,MAAM,OAAO,GAAG,GAAG,EAAE,CAAC,CAAC;QAC9D,IAAI,UAAU,CAAC,SAAS,CAAC;YAAE,OAAO,SAAS,CAAC;IAC9C,CAAC;IAED,MAAM,IAAI,KAAK,CAAC,gEAAgE,GAAG,EAAE,CAAC,CAAC;AACzF,CAAC;AAED,kFAAkF;AAElF;;;GAGG;AACH,MAAM,UAAU,kBAAkB,CAAC,aAAuB;IACxD,MAAM,QAAQ,GAAG,OAAO,CAAC,GAAG,CAAC,0BAA0B,CAAC,CAAC;IACzD,IAAI,QAAQ;QAAE,OAAO,QAAQ,CAAC;IAE9B,IAAI,aAAa,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,eAAe,CAAC;IAEvD,MAAM,MAAM,GAAG,aAAa,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACpD,OAAO,8BAA8B,MAAM,gDAAgD,CAAC;AAC9F,CAAC;AAqBD;;;;;;;;;GASG;AACH,MAAM,CAAC,KAAK,UAAU,UAAU,CAC9B,KAAa,EACb,UAA6B,EAAE;IAE/B,MAAM,MAAM,GAAG,OAAO,CAAC,SAAS,IAAI,uBAAuB,CAAC;IAC5D,SAAS,CAAC,MAAM,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAEvC,qBAAqB;IACrB,IAAI,SAAiB,CAAC;IACtB,IAAI,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC;QACjB,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,WAAW,CAAC,CAAC;QAC7C,SAAS,GAAG,MAAM,aAAa,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC;IAChD,CAAC;SAAM,CAAC;QACN,SAAS,GAAG,KAAK,CAAC;IACpB,CAAC;IAED,MAAM,IAAI,GAAG,IAAI,CAAC,QAAQ,CAAC,SAAS,EAAE,IAAI,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC;IAC/D,MAAM,cAAc,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,GAAG,IAAI,MAAM,CAAC,CAAC;IAExD,wCAAwC;IACxC,IAAI,CAAC,OAAO,CAAC,KAAK,IAAI,UAAU,CAAC,cAAc,CAAC,EAAE,CAAC;QACjD,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,cAAc,EAAE,OAAO,CAAC,CAAC;QACxD,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;QACnE,OAAO,EAAE,cAAc,EAAE,SAAS,EAAE,KAAK,CAAC,MAAM,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC;IACtE,CAAC;IAED,6BAA6B;IAC7B,MAAM,UAAU,GAAG,MAAM,mBAAmB,EAAE,CAAC;IAC/C,IAAI,CAAC,UAAU,EAAE,CAAC;QAChB,MAAM,IAAI,KAAK,CACb,8DAA8D;YAC9D,0EAA0E,CAC3E,CAAC;IACJ,CAAC;IAED,MAAM,KAAK,GAAG,QAAQ,EAAE,CAAC;IACzB,MAAM,MAAM,GAAG,OAAO,CAAC,aAAa,IAAI,eAAe,CAAC;IAExD,MAAM,IAAI,GAAG,CAAC,SAAS,EAAE,SAAS,EAAE,KAAK,EAAE,iBAAiB,EAAE,KAAK,EAAE,cAAc,EAAE,MAAM,CAAC,CAAC;IAC7F,IAAI,MAAM;QAAE,IAAI,CAAC,IAAI,CAAC,kBAAkB,EAAE,MAAM,CAAC,CAAC;IAElD,MAAM,aAAa,CAAC,UAAU,EAAE,IAAI,CAAC,CAAC;IAEtC,wEAAwE;IACxE,IAAI,CAAC,UAAU,CAAC,cAAc,CAAC,EAAE,CAAC;QAChC,+CAA+C;QAC/C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,GAAG,IAAI,CAAC,QAAQ,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;QACrE,IAAI,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;YACxB,2BAA2B;YAC3B,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;YACjD,aAAa,CAAC,cAAc,EAAE,OAAO,EAAE,OAAO,CAAC,CAAC;QAClD,CAAC;IACH,CAAC;IAED,MAAM,OAAO,GAAG,UAAU,CAAC,cAAc,CAAC;QACxC,CAAC,CAAC,MAAM,QAAQ,CAAC,cAAc,EAAE,OAAO,CAAC;QACzC,CAAC,CAAC,EAAE,CAAC;IACP,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAEnE,OAAO,EAAE,cAAc,EAAE,SAAS,EAAE,KAAK,CAAC,MAAM,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC;AACvE,CAAC;AAED,kFAAkF;AAElF;;;;;;;;GAQG;AACH,MAAM,CAAC,KAAK,UAAU,aAAa,CACjC,MAAgB,EAChB,UAA6B,EAAE;IAE/B,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAEnC,MAAM,OAAO,GAAuB,EAAE,CAAC;IACvC,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,IAAI,CAAC;YACH,OAAO,CAAC,IAAI,CAAC,MAAM,UAAU,CAAC,KAAK,EAAE,OAAO,CAAC,CAAC,CAAC;QACjD,CAAC;QAAC,OAAO,GAAG,EAAE,CAAC;YACb,OAAO,CAAC,KAAK,CAAC,qCAAqC,KAAK,KAAM,GAAa,CAAC,OAAO,EAAE,CAAC,CAAC;QACzF,CAAC;IACH,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC"}

package/dist/src/ingest/url-ingest.d.ts

@@ -0,0 +1,16 @@
+export type UrlType = 'paper' | 'github' | 'video' | 'webpage' | 'pdf' | 'image';
+export interface IngestResult {
+    id: string;
+    url: string;
+    type: UrlType;
+    title: string;
+    snippet: string;
+    fetchedAt: string;
+}
+export interface IngestOptions {
+    fetch?: typeof globalThis.fetch;
+    maxBytes?: number;
+}
+export declare function classifyUrlType(url: string): UrlType;
+export declare function ingestUrl(url: string, options?: IngestOptions): Promise<IngestResult>;
+//# sourceMappingURL=url-ingest.d.ts.map