@apmantza/greedysearch-pi 1.7.2 → 1.7.4

package/CHANGELOG.md

@@ -1,5 +1,21 @@
 # Changelog
 
+## v1.7.4 (2026-04-10)
+
+### Refactor
+- **Shared `waitForCopyButton()`** — consolidated duplicate copy-button polling loops from `bing-copilot`, `gemini`, and `coding-task` into a single `waitForCopyButton(tab, selector, { timeout, onPoll })` in `common.mjs`. Gemini's scroll-to-bottom logic passed as `onPoll` callback.
+- **Shared `TIMING` constants** — replaced 30+ scattered `setTimeout` magic numbers with named constants (`postNav`, `postNavSlow`, `postClick`, `postType`, `inputPoll`, `copyPoll`, `afterVerify`) in `common.mjs`.
+- **`waitForStreamComplete` improvements** — added `minLength` option and graceful last-value fallback; `google-ai` now uses the shared implementation instead of its own copy.
+- **Removed dead code** — deleted unused `_getOrReuseBlankTab` and `_getOrOpenEngineTab` from `bin/search.mjs`; removed unused `STREAM_POLL_INTERVAL` and `STREAM_STABLE_ROUNDS` from `coding-task`.
+
+### Fixes
+- **Synthesis tab regression** — `getOrOpenEngineTab("gemini")` call during synthesis was broken by the dead-code removal; replaced with `openNewTab()`.
+
+## v1.7.3 (2026-04-10)
+
+### Fixes
+- **Force English in Google AI results** — Added `hl=en` query parameter to Google AI Mode search URL so responses are always returned in English, regardless of the user's IP-based region (fixes #1).
+
 ## v1.7.2 (2026-04-08)
 
 ### Release
@@ -19,97 +35,97 @@
 - **README simplified** — rewritten into a shorter, concise format with quick install, usage, and layout guidance.
 
 ## v1.6.5 (2026-04-04)
-
-### Security
-- **Private URL blocking** — Added validation to block requests to localhost, RFC1918 private addresses (10.x, 192.168.x), and .local/.internal domains. Prevents accidental exposure of internal services.
-
-### Features
-- **GitHub URL rewriting** — GitHub blob URLs (`github.com/owner/repo/blob/...`) are automatically rewritten to `raw.githubusercontent.com` for faster, cleaner raw file access.
-- **GitHub repo cloning** — Root and tree URLs now trigger `git clone --depth 1` for complete repo access. Agent can explore files locally instead of parsing rendered HTML. Includes README preview and directory tree listing.
-- **Head+tail content trimming** — Large documents now use smart truncation: keeps 75% from the beginning (introduction) + 25% from the end (conclusions/examples) with `[...content trimmed...]` marker, instead of simple truncation.
-- **Anubis bot detection** — Added detection for the new Anubis proof-of-work anti-bot system (`protected by anubis`, `anubis uses a proof-of-work`).
-
-### Fixes
-- **Perplexity clipboard retry** — Added single retry with 2s delay when clipboard extraction fails, improving reliability.
-
-## v1.6.4 (2026-04-02)
-
-### Fixes
-- **Gemini scroll-to-bottom** — Changed from small random jitter scrolls to actual bottom-of-page scrolls every ~6 seconds while waiting for the copy button. This ensures lazy-loaded content is triggered and the full answer is captured.
-- **Restored missing files** — `.mjs` source files (extractors, search.mjs, launch.mjs, etc.) were incorrectly removed in v1.6.2 cleanup; now properly tracked again.
-
-## v1.6.3 (2026-04-02)
-
-### Fixes
-- **Debug output removed** — Cleaned up stderr passthrough that was causing CDP connection issues in some environments.
-
-## v1.6.2 (2026-04-01)
-
-### Fixes
-- **Anti-bot detection evasion** — Gemini synthesis now performs gentle scroll every ~6 seconds while waiting for the copy button. This prevents the button from hanging due to anti-bot "human activity" checks.
-
-## v1.6.1 (2026-03-31)
-
-### Features
-- **Single-engine full answers by default** — when using `engine: "perplexity"`, `engine: "bing"`, `engine: "google"`, or `engine: "gemini"`, the full answer is now returned by default instead of truncated previews. Multi-engine (`engine: "all"`) still uses truncated previews (~300 chars) to save tokens during synthesis. Explicit `fullAnswer: true/false` always overrides.
-
-### Code Quality
-- **Major refactoring** — extracted 438 lines from `index.ts` (856 → 418 lines) into modular formatters:
-  - `src/formatters/coding.ts` — coding task formatting
-  - `src/formatters/results.ts` — search and deep research formatting  
-  - `src/formatters/sources.ts` — source utilities (URL, label, consensus, formatting)
-  - `src/formatters/synthesis.ts` — synthesis rendering
-  - `src/utils/helpers.ts` — shared formatting utilities
-- **Complexity reduced** — cognitive complexity dropped from 360 to ~60, maintainability index improved from 11.2 to ~40+
-- **Eliminated code duplication** — removed 6 duplicate blocks, consolidated 4+ single-use helper functions
-
-### Documentation
-- Clarified `greedy_search` is WEB SEARCH ONLY — removed "NOT for codebase search" from tool description (still in skill documentation)
-
-## v1.6.0 (2026-03-29)
-
-### Breaking Changes (Backward Compatible)
-- **Merged deep_research into greedy_search** — new `depth` parameter with three levels:
-  - `fast`: single engine (~15-30s)
-  - `standard`: 3 engines + synthesis (~30-90s, default for `engine: "all"`)
-  - `deep`: 3 engines + source fetching + synthesis + confidence (~60-180s)
-- **Simpler mental model** — one tool with clear speed/quality tradeoffs instead of separate tools with overlapping flags
-- **Deprecated flags still work** — `--synthesize` maps to `depth: "standard"`, `--deep-research` maps to `depth: "deep"`
-- **deep_research tool aliased** — still works, calls `greedy_search` with `depth: "deep"`
-
-### Documentation
-- Updated README with new `depth` parameter and examples
-- Updated skill documentation (SKILL.md) to reflect simplified API
-
-## v1.5.1 (2026-03-29)
-
-- **Fixed npm package** — added `.pi-lens/` and test files to `.npmignore` to reduce package size
-
-## v1.5.0 (2026-03-29)
-
-### Features
-- **Code extraction fixed** — `coding_task` now uses clipboard interception to preserve markdown code blocks (was losing them via DOM scraping)
-- **Chrome targeting hardened** — all tools now consistently target the dedicated GreedySearch Chrome via `CDP_PROFILE_DIR`, preventing fallback to user's main Chrome session
-- **Shared utilities** — extracted ~220 lines of duplicate code from extractors into `common.mjs` (cdp wrapper, tab management, clipboard interception)
-- **Documentation leaner** — skill documentation reduced 61% (180 → 70 lines) while preserving all decision-making info
-
-### Notable
-- **NO API KEYS** — updated messaging to emphasize this works via browser automation, no API keys needed
-
-## v1.4.2 (2026-03-25)
-
-- **Fresh isolated tabs** — each search now always creates a new `about:blank` tab via `Target.createTarget` and refreshes the CDP page cache immediately after, preventing SPA navigation failures and stale DOM state from prior queries
-- **Regex-based citation extraction** — all extractors (Perplexity, Bing, Gemini) now parse sources from clipboard Markdown links (`[title](url)`) instead of DOM selectors that break on UI updates
-- **Relaxed verification detection** — `consent.mjs` now uses broad keyword matching (`includes('verify')`, `includes('human')`) instead of anchored regexes, correctly catching button text variants like "Verify you are human" across Cloudflare, Microsoft, and generic modals
-
-## v1.4.1
-
-- **Fixed parallel synthesis** — multiple `greedy_search` calls with `synthesize: true` now run safely in parallel. Each search creates a fresh Gemini tab that gets cleaned up after synthesis, preventing tab conflicts and "Uncaught" errors.
-
-## v1.4.0
-
-- **Grounded synthesis** — Gemini now receives a normalized source registry with stable source IDs, agreement summaries, caveats, and cited claims
-- **Real deep research** — top sources are fetched before synthesis so deep research answers are grounded in fetched evidence, not just engine summaries
-- **Richer source metadata** — source output now includes canonical URLs, domains, source types, per-engine attribution, and confidence metadata
-- **Cleaner tab lifecycle** — temporary Perplexity, Bing, and Google tabs are closed after each fan-out search, and synthesis finishes on the Gemini tab
-- **Isolated Chrome targeting** — GreedySearch now refuses to fall back to your normal Chrome session, preventing stray remote-debugging prompts
+
+### Security
+- **Private URL blocking** — Added validation to block requests to localhost, RFC1918 private addresses (10.x, 192.168.x), and .local/.internal domains. Prevents accidental exposure of internal services.
+
+### Features
+- **GitHub URL rewriting** — GitHub blob URLs (`github.com/owner/repo/blob/...`) are automatically rewritten to `raw.githubusercontent.com` for faster, cleaner raw file access.
+- **GitHub repo cloning** — Root and tree URLs now trigger `git clone --depth 1` for complete repo access. Agent can explore files locally instead of parsing rendered HTML. Includes README preview and directory tree listing.
+- **Head+tail content trimming** — Large documents now use smart truncation: keeps 75% from the beginning (introduction) + 25% from the end (conclusions/examples) with `[...content trimmed...]` marker, instead of simple truncation.
+- **Anubis bot detection** — Added detection for the new Anubis proof-of-work anti-bot system (`protected by anubis`, `anubis uses a proof-of-work`).
+
+### Fixes
+- **Perplexity clipboard retry** — Added single retry with 2s delay when clipboard extraction fails, improving reliability.
+
+## v1.6.4 (2026-04-02)
+
+### Fixes
+- **Gemini scroll-to-bottom** — Changed from small random jitter scrolls to actual bottom-of-page scrolls every ~6 seconds while waiting for the copy button. This ensures lazy-loaded content is triggered and the full answer is captured.
+- **Restored missing files** — `.mjs` source files (extractors, search.mjs, launch.mjs, etc.) were incorrectly removed in v1.6.2 cleanup; now properly tracked again.
+
+## v1.6.3 (2026-04-02)
+
+### Fixes
+- **Debug output removed** — Cleaned up stderr passthrough that was causing CDP connection issues in some environments.
+
+## v1.6.2 (2026-04-01)
+
+### Fixes
+- **Anti-bot detection evasion** — Gemini synthesis now performs gentle scroll every ~6 seconds while waiting for the copy button. This prevents the button from hanging due to anti-bot "human activity" checks.
+
+## v1.6.1 (2026-03-31)
+
+### Features
+- **Single-engine full answers by default** — when using `engine: "perplexity"`, `engine: "bing"`, `engine: "google"`, or `engine: "gemini"`, the full answer is now returned by default instead of truncated previews. Multi-engine (`engine: "all"`) still uses truncated previews (~300 chars) to save tokens during synthesis. Explicit `fullAnswer: true/false` always overrides.
+
+### Code Quality
+- **Major refactoring** — extracted 438 lines from `index.ts` (856 → 418 lines) into modular formatters:
+  - `src/formatters/coding.ts` — coding task formatting
+  - `src/formatters/results.ts` — search and deep research formatting  
+  - `src/formatters/sources.ts` — source utilities (URL, label, consensus, formatting)
+  - `src/formatters/synthesis.ts` — synthesis rendering
+  - `src/utils/helpers.ts` — shared formatting utilities
+- **Complexity reduced** — cognitive complexity dropped from 360 to ~60, maintainability index improved from 11.2 to ~40+
+- **Eliminated code duplication** — removed 6 duplicate blocks, consolidated 4+ single-use helper functions
+
+### Documentation
+- Clarified `greedy_search` is WEB SEARCH ONLY — removed "NOT for codebase search" from tool description (still in skill documentation)
+
+## v1.6.0 (2026-03-29)
+
+### Breaking Changes (Backward Compatible)
+- **Merged deep_research into greedy_search** — new `depth` parameter with three levels:
+  - `fast`: single engine (~15-30s)
+  - `standard`: 3 engines + synthesis (~30-90s, default for `engine: "all"`)
+  - `deep`: 3 engines + source fetching + synthesis + confidence (~60-180s)
+- **Simpler mental model** — one tool with clear speed/quality tradeoffs instead of separate tools with overlapping flags
+- **Deprecated flags still work** — `--synthesize` maps to `depth: "standard"`, `--deep-research` maps to `depth: "deep"`
+- **deep_research tool aliased** — still works, calls `greedy_search` with `depth: "deep"`
+
+### Documentation
+- Updated README with new `depth` parameter and examples
+- Updated skill documentation (SKILL.md) to reflect simplified API
+
+## v1.5.1 (2026-03-29)
+
+- **Fixed npm package** — added `.pi-lens/` and test files to `.npmignore` to reduce package size
+
+## v1.5.0 (2026-03-29)
+
+### Features
+- **Code extraction fixed** — `coding_task` now uses clipboard interception to preserve markdown code blocks (was losing them via DOM scraping)
+- **Chrome targeting hardened** — all tools now consistently target the dedicated GreedySearch Chrome via `CDP_PROFILE_DIR`, preventing fallback to user's main Chrome session
+- **Shared utilities** — extracted ~220 lines of duplicate code from extractors into `common.mjs` (cdp wrapper, tab management, clipboard interception)
+- **Documentation leaner** — skill documentation reduced 61% (180 → 70 lines) while preserving all decision-making info
+
+### Notable
+- **NO API KEYS** — updated messaging to emphasize this works via browser automation, no API keys needed
+
+## v1.4.2 (2026-03-25)
+
+- **Fresh isolated tabs** — each search now always creates a new `about:blank` tab via `Target.createTarget` and refreshes the CDP page cache immediately after, preventing SPA navigation failures and stale DOM state from prior queries
+- **Regex-based citation extraction** — all extractors (Perplexity, Bing, Gemini) now parse sources from clipboard Markdown links (`[title](url)`) instead of DOM selectors that break on UI updates
+- **Relaxed verification detection** — `consent.mjs` now uses broad keyword matching (`includes('verify')`, `includes('human')`) instead of anchored regexes, correctly catching button text variants like "Verify you are human" across Cloudflare, Microsoft, and generic modals
+
+## v1.4.1
+
+- **Fixed parallel synthesis** — multiple `greedy_search` calls with `synthesize: true` now run safely in parallel. Each search creates a fresh Gemini tab that gets cleaned up after synthesis, preventing tab conflicts and "Uncaught" errors.
+
+## v1.4.0
+
+- **Grounded synthesis** — Gemini now receives a normalized source registry with stable source IDs, agreement summaries, caveats, and cited claims
+- **Real deep research** — top sources are fetched before synthesis so deep research answers are grounded in fetched evidence, not just engine summaries
+- **Richer source metadata** — source output now includes canonical URLs, domains, source types, per-engine attribution, and confidence metadata
+- **Cleaner tab lifecycle** — temporary Perplexity, Bing, and Google tabs are closed after each fan-out search, and synthesis finishes on the Gemini tab
+- **Isolated Chrome targeting** — GreedySearch now refuses to fall back to your normal Chrome session, preventing stray remote-debugging prompts

package/README.md

@@ -1,73 +1,73 @@
-# GreedySearch for Pi
-
-Multi-engine AI web search for Pi via browser automation.
-
-- No API keys
-- Real browser results (Perplexity, Bing Copilot, Google AI)
-- Optional Gemini synthesis with source grounding
-
-## Install
-
-```bash
-pi install npm:@apmantza/greedysearch-pi
-```
-
-Or from git:
-
-```bash
-pi install git:github.com/apmantza/GreedySearch-pi
-```
-
-## Tools
-
-- `greedy_search` - fast or grounded multi-engine search
-- `coding_task` - browser-routed Gemini/Copilot coding assistance
-
-## Quick usage
-
-```js
-greedy_search({ query: "React 19 changes" })
-greedy_search({ query: "Prisma vs Drizzle", engine: "all", depth: "fast" })
-greedy_search({ query: "Best auth architecture 2026", engine: "all", depth: "deep" })
-```
-
-## Parameters (`greedy_search`)
-
-- `query` (required)
-- `engine`: `all` (default), `perplexity`, `bing`, `google`, `gemini`
-- `depth`: `standard` (default), `fast`, `deep`
-- `fullAnswer`: return full single-engine output instead of preview
-
-## Depth modes
-
-- `fast` - quickest, no synthesis/source fetching
-- `standard` - balanced default for `engine: "all"` (synthesis + fetched sources)
-- `deep` - strongest grounding and confidence metadata
-
-## Runtime commands
-
-```bash
-node ~/.pi/agent/git/GreedySearch-pi/bin/launch.mjs
-node ~/.pi/agent/git/GreedySearch-pi/bin/launch.mjs --status
-node ~/.pi/agent/git/GreedySearch-pi/bin/launch.mjs --kill
-```
-
-## Requirements
-
-- Chrome
-- Node.js 22+
-
-## Project layout
-
-- `bin/` - runtime CLIs (`search.mjs`, `launch.mjs`, `cdp.mjs`, `coding-task.mjs`)
-- `extractors/` - engine-specific automation
-- `src/` - ranking/fetching/formatting internals
-- `skills/` - Pi skill metadata
-
-## Changelog
-
-See `CHANGELOG.md`.
-
-## License
-
-MIT
+# GreedySearch for Pi
+
+Multi-engine AI web search for Pi via browser automation.
+
+- No API keys
+- Real browser results (Perplexity, Bing Copilot, Google AI)
+- Optional Gemini synthesis with source grounding
+
+## Install
+
+```bash
+pi install npm:@apmantza/greedysearch-pi
+```
+
+Or from git:
+
+```bash
+pi install git:github.com/apmantza/GreedySearch-pi
+```
+
+## Tools
+
+- `greedy_search` - fast or grounded multi-engine search
+- `coding_task` - browser-routed Gemini/Copilot coding assistance
+
+## Quick usage
+
+```js
+greedy_search({ query: "React 19 changes" })
+greedy_search({ query: "Prisma vs Drizzle", engine: "all", depth: "fast" })
+greedy_search({ query: "Best auth architecture 2026", engine: "all", depth: "deep" })
+```
+
+## Parameters (`greedy_search`)
+
+- `query` (required)
+- `engine`: `all` (default), `perplexity`, `bing`, `google`, `gemini`
+- `depth`: `standard` (default), `fast`, `deep`
+- `fullAnswer`: return full single-engine output instead of preview
+
+## Depth modes
+
+- `fast` - quickest, no synthesis/source fetching
+- `standard` - balanced default for `engine: "all"` (synthesis + fetched sources)
+- `deep` - strongest grounding and confidence metadata
+
+## Runtime commands
+
+```bash
+node ~/.pi/agent/git/GreedySearch-pi/bin/launch.mjs
+node ~/.pi/agent/git/GreedySearch-pi/bin/launch.mjs --status
+node ~/.pi/agent/git/GreedySearch-pi/bin/launch.mjs --kill
+```
+
+## Requirements
+
+- Chrome
+- Node.js 22+
+
+## Project layout
+
+- `bin/` - runtime CLIs (`search.mjs`, `launch.mjs`, `cdp.mjs`, `coding-task.mjs`)
+- `extractors/` - engine-specific automation
+- `src/` - ranking/fetching/formatting internals
+- `skills/` - Pi skill metadata
+
+## Changelog
+
+See `CHANGELOG.md`.
+
+## License
+
+MIT

package/bin/coding-task.mjs

@@ -12,8 +12,8 @@
 import { existsSync, readFileSync, writeFileSync } from "node:fs";
 import { tmpdir } from "node:os";
 import { fileURLToPath } from "node:url";
-import { cdp, injectClipboardInterceptor } from "../extractors/common.mjs";
-import { dismissConsent, handleVerification } from "../extractors/consent.mjs";
+import { cdp, injectClipboardInterceptor, waitForCopyButton } from "../extractors/common.mjs";
+import { dismissConsent, handleVerification } from "../extractors/consent.mjs";
 
 const __dir = fileURLToPath(new URL(".", import.meta.url));
 const PAGES_CACHE = `${tmpdir().replace(/\\/g, "/")}/cdp-pages.json`;
@@ -31,9 +31,7 @@ const MODE_PROMPTS = {
 	debug: `You are a senior engineer debugging someone else's code. You have fresh eyes — no prior assumptions about what should work. Given the bug description and relevant code: (1) identify the most likely root cause, being specific about the exact line or condition, (2) explain why it manifests the way it does, (3) suggest the minimal fix, (4) flag any other latent bugs you notice while reading. Do not guess vaguely — reason from the code. If you need information that isn't provided, say exactly what you'd add to narrow it down.`,
 };
 
-const STREAM_POLL_INTERVAL = 800;
-const STREAM_STABLE_ROUNDS = 4;
-const STREAM_TIMEOUT = 120000; // coding tasks take longer
+const STREAM_TIMEOUT = 120000; // coding tasks take longer — passed to waitForCopyButton
 const MIN_RESPONSE_LENGTH = 50;
 
 // ---------------------------------------------------------------------------
@@ -104,17 +102,7 @@ const ENGINES = {
 		},
 
 		async waitForCopyButton(tab) {
-			const deadline = Date.now() + STREAM_TIMEOUT;
-			while (Date.now() < deadline) {
-				await new Promise((r) => setTimeout(r, STREAM_POLL_INTERVAL));
-				const found = await cdp([
-					"eval",
-					tab,
-					`!!document.querySelector('button[aria-label="Copy"]')`,
-				]).catch(() => "false");
-				if (found === "true") return;
-			}
-			throw new Error("Gemini copy button did not appear within timeout");
+			await waitForCopyButton(tab, 'button[aria-label="Copy"]', { timeout: STREAM_TIMEOUT });
 		},
 
 		async extract(tab) {
@@ -162,17 +150,7 @@ const ENGINES = {
 		},
 
 		async waitForCopyButton(tab) {
-			const deadline = Date.now() + STREAM_TIMEOUT;
-			while (Date.now() < deadline) {
-				await new Promise((r) => setTimeout(r, STREAM_POLL_INTERVAL));
-				const found = await cdp([
-					"eval",
-					tab,
-					`!!document.querySelector('button[data-testid="copy-ai-message-button"]')`,
-				]).catch(() => "false");
-				if (found === "true") return;
-			}
-			throw new Error("Copilot copy button did not appear within timeout");
+			await waitForCopyButton(tab, 'button[data-testid="copy-ai-message-button"]', { timeout: STREAM_TIMEOUT });
 		},
 
 		async extract(tab) {