@apmantza/greedysearch-pi 1.9.0 → 1.9.2

package/extractors/selectors.mjs

@@ -1,54 +1,55 @@
-// extractors/selectors.mjs
-// Centralized CSS selectors for all engines.
-// Update selectors here when a site changes its UI.
-
-export const SELECTORS = {
-	// ──────────────────────────────────────────────
-	// Perplexity (perplexity.ai)
-	// ──────────────────────────────────────────────
-	perplexity: {
-		input: "#ask-input",
-		// Note: copy button found via JS in extractor (language-agnostic)
-		copyButton: null,
-		sourceItem: "[data-pplx-citation-url]",
-		sourceLink: "a",
-		consent: "#onetrust-accept-btn-handler",
-	},
-
-	// ──────────────────────────────────────────────
-	// Bing Copilot (copilot.microsoft.com)
-	// ──────────────────────────────────────────────
-	bing: {
-		input: "#userInput",
-		copyButton: 'button[data-testid="copy-ai-message-button"]',
-		sourceLink: 'a[href^="http"][target="_blank"]',
-		sourceExclude: "copilot.microsoft.com",
-		consent: "#onetrust-accept-btn-handler",
-	},
-
-	// ──────────────────────────────────────────────
-	// Google AI Mode (google.com/search?udm=50)
-	// ──────────────────────────────────────────────
-	google: {
-		answerContainer: ".pWvJNd",
-		sourceLink: 'a[href^="http"]',
-		sourceExclude: ["google.", "gstatic", "googleapis"],
-		sourceHeadingParent: "[data-snhf]",
-		consent: '#L2AGLb, button[jsname="b3VHJd"], .tHlp8d',
-	},
-
-	// ──────────────────────────────────────────────
-	// Gemini (gemini.google.com/app)
-	// ──────────────────────────────────────────────
-	gemini: {
-		input: "rich-textarea .ql-editor",
-		// Language-agnostic: use Material icon data attributes (work across locales)
-		copyButton: 'button:has(mat-icon[data-mat-icon-name="content_copy"])',
-		sendButton: 'button:has(mat-icon[data-mat-icon-name="send"]), .send-button',
-		sourcesSidebarButton: "button.legacy-sources-sidebar-button",
-		sourcesExclude: ["gemini.google", "gstatic", "google.com/search"],
-		citationButtonPattern: 'button[aria-label*="citation from"]',
-		// For parsing citation aria-labels: "View source details for citation from {name}. Opens side panel."
-		citationNameRegex: /from\s+([^.]+)\.\s/,
-	},
-};
+// extractors/selectors.mjs
+// Centralized CSS selectors for all engines.
+// Update selectors here when a site changes its UI.
+
+export const SELECTORS = {
+	// ──────────────────────────────────────────────
+	// Perplexity (perplexity.ai)
+	// ──────────────────────────────────────────────
+	perplexity: {
+		input: "#ask-input",
+		// Note: copy button found via JS in extractor (language-agnostic)
+		copyButton: null,
+		sourceItem: "[data-pplx-citation-url]",
+		sourceLink: "a",
+		consent: "#onetrust-accept-btn-handler",
+	},
+
+	// ──────────────────────────────────────────────
+	// Bing Copilot (copilot.microsoft.com)
+	// ──────────────────────────────────────────────
+	bing: {
+		input: "#userInput",
+		copyButton: 'button[data-testid="copy-ai-message-button"]',
+		sourceLink: 'a[href^="http"][target="_blank"]',
+		sourceExclude: "copilot.microsoft.com",
+		consent: "#onetrust-accept-btn-handler",
+	},
+
+	// ──────────────────────────────────────────────
+	// Google AI Mode (google.com/search?udm=50)
+	// ──────────────────────────────────────────────
+	google: {
+		answerContainer: ".pWvJNd",
+		sourceLink: 'a[href^="http"]',
+		sourceExclude: ["google.", "gstatic", "googleapis"],
+		sourceHeadingParent: "[data-snhf]",
+		consent: '#L2AGLb, button[jsname="b3VHJd"], .tHlp8d',
+	},
+
+	// ──────────────────────────────────────────────
+	// Gemini (gemini.google.com/app)
+	// ──────────────────────────────────────────────
+	gemini: {
+		input: "rich-textarea .ql-editor",
+		// Language-agnostic: use Material icon data attributes (work across locales)
+		copyButton: 'button:has(mat-icon[data-mat-icon-name="content_copy"])',
+		sendButton: 'button:has(mat-icon[data-mat-icon-name="send"]), .send-button',
+		sourcesSidebarButton: "button.legacy-sources-sidebar-button",
+		sourcesExclude: ["gemini.google", "gstatic", "google.com/search"],
+		citationButtonPattern: 'button[aria-label*="citation from"]',
+		// For parsing citation aria-labels: "View source details for citation from {name}. Opens side panel."
+		// Bounded + non-overlapping character classes to prevent ReDoS
+		citationNameRegex: /from\s{1,20}([^.]{1,200})\.\s/,
+	},
+};

package/index.ts

@@ -1,177 +1,176 @@
-/**
- * GreedySearch Pi Extension
- *
- * Adds `greedy_search` tool to Pi.
- * Use depth: "deep" for deep research (source fetching + synthesis + confidence).
- *
- * Reports streaming progress as each engine completes.
- * Requires Chrome to be running (or it auto-launches a dedicated instance).
- */
-
-import { spawn } from "node:child_process";
-import { dirname, join } from "node:path";
-import { fileURLToPath } from "node:url";
-import type { ExtensionAPI } from "@mariozechner/pi-coding-agent";
-
-import { registerGreedySearchTool } from "./src/tools/greedy-search-handler.js";
-import { cdpAvailable } from "./src/tools/shared.js";
-
-const __dir = dirname(fileURLToPath(import.meta.url));
-
-export default function greedySearchExtension(pi: ExtensionAPI) {
-	pi.on("session_start", async (_event, ctx) => {
-		if (!cdpAvailable(__dir)) {
-			ctx.ui.notify(
-				"GreedySearch: cdp.mjs missing from package directory — try reinstalling: pi install git:github.com/apmantza/GreedySearch-pi",
-				"warning",
-			);
-		}
-	});
-
-	// ─── greedy_search ────────────────────────────────────────────────────────
-	registerGreedySearchTool(pi, __dir);
-
-	// ─── GreedySearch Chrome commands ─────────────────────────────────────────
-	pi.registerCommand("greedy-visible", {
-		description:
-			"Launch GreedySearch Chrome in visible mode for captcha/login/cookie setup.",
-		handler: async (_args, ctx) => {
-			await runChromeCommand([], ctx, "Visible GreedySearch Chrome launched.");
-		},
-	});
-
-	pi.registerCommand("greedy-status", {
-		description: "Show GreedySearch Chrome status.",
-		handler: async (_args, ctx) => {
-			await runChromeCommand(["--status"], ctx);
-		},
-	});
-
-	pi.registerCommand("greedy-kill", {
-		description: "Stop GreedySearch Chrome.",
-		handler: async (_args, ctx) => {
-			await runChromeCommand(["--kill"], ctx, "GreedySearch Chrome stopped.");
-		},
-	});
-
-	// ─── /set-greedy-locale command ───────────────────────────────────────────
-	pi.registerCommand("set-greedy-locale", {
-		description:
-			"Set default locale for GreedySearch results (e.g., /set-greedy-locale de, /set-greedy-locale --clear, /set-greedy-locale --show)",
-		handler: async (args, ctx) => {
-			const arg = args.trim() || "--show";
-
-			if (arg === "--show") {
-				const config = loadUserConfig();
-				if (config.locale) {
-					ctx.ui.notify(`Default locale: ${config.locale}`, "info");
-				} else {
-					ctx.ui.notify("No default locale (uses: en)", "info");
-				}
-				return;
-			}
-
-			if (arg === "--clear") {
-				const config = loadUserConfig();
-				delete config.locale;
-				saveUserConfig(config);
-				ctx.ui.notify("Default locale cleared (now uses: en).", "info");
-				return;
-			}
-
-			// Set locale
-			const locale = arg.toLowerCase();
-			const VALID_LOCALES = [
-				"en",
-				"de",
-				"fr",
-				"es",
-				"it",
-				"pt",
-				"nl",
-				"pl",
-				"ru",
-				"ja",
-				"ko",
-				"zh",
-				"ar",
-				"hi",
-				"tr",
-				"sv",
-				"da",
-				"no",
-				"fi",
-				"cs",
-				"hu",
-				"ro",
-				"el",
-			];
-
-			if (!VALID_LOCALES.includes(locale)) {
-				ctx.ui.notify(
-					`Invalid locale "${locale}". Valid: ${VALID_LOCALES.join(", ")}`,
-					"error",
-				);
-				return;
-			}
-
-			const config = loadUserConfig();
-			config.locale = locale;
-			saveUserConfig(config);
-			ctx.ui.notify(`Default locale set to: ${locale}`, "info");
-		},
-	});
-}
-
-import { existsSync, mkdirSync, readFileSync, writeFileSync } from "node:fs";
-// Config helpers for /set-greedy-locale command
-import { homedir } from "node:os";
-
-const USER_CONFIG_DIR = join(homedir(), ".config", "greedysearch");
-const USER_CONFIG_FILE = join(USER_CONFIG_DIR, "config.json");
-
-async function runChromeCommand(
-	args: string[],
-	ctx: any,
-	successMessage?: string,
-): Promise<void> {
-	const visibleBin = join(__dir, "bin", "visible.mjs");
-	const { code, output } = await new Promise<{
-		code: number | null;
-		output: string;
-	}>((resolve) => {
-		const proc = spawn(process.execPath, [visibleBin, ...args], {
-			stdio: ["ignore", "pipe", "pipe"],
-			env: { ...process.env, GREEDY_SEARCH_VISIBLE: "1" },
-		});
-		let output = "";
-		proc.stdout.on("data", (d: Buffer) => (output += d.toString()));
-		proc.stderr.on("data", (d: Buffer) => (output += d.toString()));
-		proc.on("close", (code: number | null) => resolve({ code, output }));
-	});
-
-	if (code === 0) {
-		ctx.ui.notify((successMessage || output.trim() || "Done.").trim(), "info");
-	} else {
-		ctx.ui.notify(
-			output.trim() || `GreedySearch Chrome command failed (${code})`,
-			"error",
-		);
-	}
-}
-
-function loadUserConfig(): Record<string, string> {
-	try {
-		if (existsSync(USER_CONFIG_FILE)) {
-			return JSON.parse(readFileSync(USER_CONFIG_FILE, "utf8"));
-		}
-	} catch {
-		// Ignore parse errors
-	}
-	return {};
-}
-
-function saveUserConfig(config: Record<string, string>): void {
-	mkdirSync(USER_CONFIG_DIR, { recursive: true });
-	writeFileSync(USER_CONFIG_FILE, JSON.stringify(config, null, 2), "utf8");
-}
+/**
+ * GreedySearch Pi Extension
+ *
+ * Adds `greedy_search` tool to Pi.
+ * Use depth: "deep" for deep research (source fetching + synthesis + confidence).
+ *
+ * Reports streaming progress as each engine completes.
+ * Requires Chrome to be running (or it auto-launches a dedicated instance).
+ */
+
+import { spawn } from "node:child_process";
+import { existsSync, mkdirSync, readFileSync, writeFileSync } from "node:fs";
+import { homedir } from "node:os";
+import { dirname, join } from "node:path";
+import { fileURLToPath } from "node:url";
+
+import { registerGreedySearchTool } from "./src/tools/greedy-search-handler.js";
+
+type ExtensionAPI = any;
+import { cdpAvailable } from "./src/tools/shared.js";
+
+const __dir = dirname(fileURLToPath(import.meta.url));
+
+export default function greedySearchExtension(pi: ExtensionAPI) {
+	pi.on("session_start", async (_event, ctx) => {
+		if (!cdpAvailable(__dir)) {
+			ctx.ui.notify(
+				"GreedySearch: cdp.mjs missing from package directory — try reinstalling: pi install git:github.com/apmantza/GreedySearch-pi",
+				"warning",
+			);
+		}
+	});
+
+	// ─── greedy_search ────────────────────────────────────────────────────────
+	registerGreedySearchTool(pi, __dir);
+
+	// ─── GreedySearch Chrome commands ─────────────────────────────────────────
+	pi.registerCommand("greedy-visible", {
+		description:
+			"Launch GreedySearch Chrome in visible mode for captcha/login/cookie setup.",
+		handler: async (_args, ctx) => {
+			await runChromeCommand([], ctx, "Visible GreedySearch Chrome launched.");
+		},
+	});
+
+	pi.registerCommand("greedy-status", {
+		description: "Show GreedySearch Chrome status.",
+		handler: async (_args, ctx) => {
+			await runChromeCommand(["--status"], ctx);
+		},
+	});
+
+	pi.registerCommand("greedy-kill", {
+		description: "Stop GreedySearch Chrome.",
+		handler: async (_args, ctx) => {
+			await runChromeCommand(["--kill"], ctx, "GreedySearch Chrome stopped.");
+		},
+	});
+
+	// ─── /set-greedy-locale command ───────────────────────────────────────────
+	pi.registerCommand("set-greedy-locale", {
+		description:
+			"Set default locale for GreedySearch results (e.g., /set-greedy-locale de, /set-greedy-locale --clear, /set-greedy-locale --show)",
+		handler: async (args, ctx) => {
+			const arg = args.trim() || "--show";
+
+			if (arg === "--show") {
+				const config = loadUserConfig();
+				if (config.locale) {
+					ctx.ui.notify(`Default locale: ${config.locale}`, "info");
+				} else {
+					ctx.ui.notify("No default locale (uses: en)", "info");
+				}
+				return;
+			}
+
+			if (arg === "--clear") {
+				const config = loadUserConfig();
+				delete config.locale;
+				saveUserConfig(config);
+				ctx.ui.notify("Default locale cleared (now uses: en).", "info");
+				return;
+			}
+
+			// Set locale
+			const locale = arg.toLowerCase();
+			const VALID_LOCALES = [
+				"en",
+				"de",
+				"fr",
+				"es",
+				"it",
+				"pt",
+				"nl",
+				"pl",
+				"ru",
+				"ja",
+				"ko",
+				"zh",
+				"ar",
+				"hi",
+				"tr",
+				"sv",
+				"da",
+				"no",
+				"fi",
+				"cs",
+				"hu",
+				"ro",
+				"el",
+			];
+
+			if (!VALID_LOCALES.includes(locale)) {
+				ctx.ui.notify(
+					`Invalid locale "${locale}". Valid: ${VALID_LOCALES.join(", ")}`,
+					"error",
+				);
+				return;
+			}
+
+			const config = loadUserConfig();
+			config.locale = locale;
+			saveUserConfig(config);
+			ctx.ui.notify(`Default locale set to: ${locale}`, "info");
+		},
+	});
+}
+
+const USER_CONFIG_DIR = join(homedir(), ".config", "greedysearch");
+const USER_CONFIG_FILE = join(USER_CONFIG_DIR, "config.json");
+
+async function runChromeCommand(
+	args: string[],
+	ctx: any,
+	successMessage?: string,
+): Promise<void> {
+	const visibleBin = join(__dir, "bin", "visible.mjs");
+	const { code, output } = await new Promise<{
+		code: number | null;
+		output: string;
+	}>((resolve) => {
+		const proc = spawn(process.execPath, [visibleBin, ...args], {
+			stdio: ["ignore", "pipe", "pipe"],
+			env: { ...process.env, GREEDY_SEARCH_VISIBLE: "1" },
+		});
+		let output = "";
+		proc.stdout.on("data", (d: Buffer) => (output += d.toString()));
+		proc.stderr.on("data", (d: Buffer) => (output += d.toString()));
+		proc.on("close", (code: number | null) => resolve({ code, output }));
+	});
+
+	if (code === 0) {
+		ctx.ui.notify((successMessage || output.trim() || "Done.").trim(), "info");
+	} else {
+		ctx.ui.notify(
+			output.trim() || `GreedySearch Chrome command failed (${code})`,
+			"error",
+		);
+	}
+}
+
+function loadUserConfig(): Record<string, string> {
+	try {
+		if (existsSync(USER_CONFIG_FILE)) {
+			return JSON.parse(readFileSync(USER_CONFIG_FILE, "utf8"));
+		}
+	} catch {
+		// Ignore parse errors
+	}
+	return {};
+}
+
+function saveUserConfig(config: Record<string, string>): void {
+	mkdirSync(USER_CONFIG_DIR, { recursive: true });
+	writeFileSync(USER_CONFIG_FILE, JSON.stringify(config, null, 2), "utf8");
+}

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@apmantza/greedysearch-pi",
-	"version": "1.9.0",
+	"version": "1.9.2",
 	"description": "Headless multi-engine AI search (Perplexity, Bing Copilot, Google AI) via browser automation -- NO API KEYS needed. Extracts answers with sources, optional synthesis. Grounded AI answers from real browser interactions.",
 	"type": "module",
 	"keywords": [
@@ -45,11 +45,16 @@
 	},
 	"dependencies": {
 		"@mozilla/readability": "^0.6.0",
+		"@sinclair/typebox": "^0.34.48",
 		"jsdom": "^24.0.0",
 		"turndown": "^7.1.2"
 	},
 	"peerDependencies": {
-		"@mariozechner/pi-coding-agent": "*",
-		"@sinclair/typebox": "*"
+		"@earendil-works/pi-coding-agent": "*"
+	},
+	"peerDependenciesMeta": {
+		"@earendil-works/pi-coding-agent": {
+			"optional": true
+		}
 	}
 }

package/skills/greedy-search/skill.md

@@ -1,26 +1,12 @@
 ---
 name: greedy-search
-description: Live web search via Perplexity, Bing, Google AI, and Gemini. Use for current docs, recent errors/framework changes, dependency choices, or stale-knowledge questions. NOT for codebase search.
+description: Web search via Perplexity, Bing, Google AI & Gemini. Current docs, recent changes, dependency choices. NOT codebase search.
 ---
 
-Use `greedy_search` for live web answers.
+`greedy_search({ query, engine: "all"|"perplexity"|"bing"|"google"|"gemini", depth: "fast"|"standard"|"deep"|"research", breadth: 1-5, iterations: 1-3, maxSources: 3-12, visible: bool })`
 
-```js
-greedy_search({ query: "React 19 changes", depth: "standard" });
-```
+**Depth:** `fast`(15-30s, no synthesis) · `standard`(30-90s, all+synthesis+sources) · `deep`(60-180s, stronger grounding) · `research`(slowest, iterative planning+follow-ups+learning extraction; uses breadth/iterations/maxSources)
 
-**Params:** `query` (required), `engine`: `all`|`perplexity`|`bing`|`google`|`gemini`, `depth`: `fast`|`standard`|`deep`, `fullAnswer`, `visible`/`alwaysVisible`/`headless: false`
+**Auto-recovery:** Headless default. Bing/Perplexity auto-retry visible on CF block. Manual CAPTCHA → visible stays open; solve then rerun.
 
-**Depths:**
-
-- `fast`: ~15-30s, single engine, no synthesis
-- `standard`: ~30-90s, all engines + Gemini synthesis + sources
-- `deep`: ~60-180s, stronger grounding + confidence metadata
-
-**Captcha/blocks:** Headless by default. Bing/Perplexity auto-retry in visible mode when blocked. If human verification is needed, visible Chrome stays open — tell the user to solve it and rerun. Use `visible: true` proactively for repeated issues.
-
-**Pi commands:** `/greedy-visible`, `/greedy-status`, `/greedy-kill`, `/set-greedy-locale`
-
-**CDP safety:** Never call raw `bin/cdp.mjs`. Use `bin/cdp-greedy.mjs`, `bin/cdp-visible.mjs`, or `bin/cdp-headless.mjs`.
-
-Old `coding_task`/`deep_research` folded into `greedy_search`. Use `engine: "gemini"` for one-off opinion, `depth: "deep"` for research.
+**CDP safety:** Use `bin/cdp-greedy.mjs` only. Never raw `bin/cdp.mjs`.