gsd-pi 0.1.0

package/src/resources/extensions/search-the-web/tool-search.ts

@@ -0,0 +1,424 @@
+/**
+ * search-the-web tool — Rich web search with full Brave API support.
+ *
+ * Improvements over v1:
+ * - Extra snippets (up to 5 per result) for 3-5x more useful context
+ * - Freshness filtering (day/week/month/year/custom)
+ * - Domain filtering
+ * - AI summarizer integration (Brave Summarizer API)
+ * - Result age/date surfaced
+ * - Auto-freshness detection for recency-sensitive queries
+ * - Token-efficient compact output format
+ * - Richer parameter set for precise searching
+ */
+
+import type { ExtensionAPI } from "@mariozechner/pi-coding-agent";
+import { truncateHead, formatSize, DEFAULT_MAX_BYTES, DEFAULT_MAX_LINES } from "@mariozechner/pi-coding-agent";
+import { Text } from "@mariozechner/pi-tui";
+import { Type } from "@sinclair/typebox";
+import { StringEnum } from "@mariozechner/pi-ai";
+
+import { LRUTTLCache } from "./cache";
+import { fetchWithRetry, HttpError } from "./http";
+import { normalizeQuery, toDedupeKey, detectFreshness } from "./url-utils";
+import { formatSearchResults, type SearchResultFormatted } from "./format";
+
+// =============================================================================
+// Types
+// =============================================================================
+
+interface BraveWebResult {
+  title: string;
+  url: string;
+  description: string;
+  age?: string;
+  page_age?: string;
+  language?: string;
+  extra_snippets?: string[];
+  meta_url?: { scheme?: string; netloc?: string; hostname?: string; path?: string };
+  [key: string]: unknown;
+}
+
+interface BraveSummarizerResponse {
+  type?: string;
+  status?: number;
+  title?: string;
+  summary?: Array<{ type: string; data: string }>;
+  enrichments?: unknown;
+  [key: string]: unknown;
+}
+
+interface CachedSearchResult {
+  results: SearchResultFormatted[];
+  summarizerKey?: string;
+}
+
+// =============================================================================
+// Caches
+// =============================================================================
+
+// Search results: max 100 entries, 10-minute TTL
+const searchCache = new LRUTTLCache<CachedSearchResult>({ max: 100, ttlMs: 600_000 });
+searchCache.startPurgeInterval(60_000);
+
+// Summarizer responses: max 50 entries, 15-minute TTL
+const summarizerCache = new LRUTTLCache<string>({ max: 50, ttlMs: 900_000 });
+
+// =============================================================================
+// Brave API helpers
+// =============================================================================
+
+function getBraveApiKey(): string {
+  return process.env.BRAVE_API_KEY || "";
+}
+
+function braveHeaders(): Record<string, string> {
+  return {
+    "Accept": "application/json",
+    "Accept-Encoding": "gzip",
+    "X-Subscription-Token": getBraveApiKey(),
+  };
+}
+
+/**
+ * Normalize a Brave result into our formatted result type.
+ * Extracts the useful fields and discards the rest.
+ */
+function normalizeBraveResult(r: BraveWebResult): SearchResultFormatted {
+  return {
+    title: r.title || "(untitled)",
+    url: r.url,
+    description: r.description || "",
+    age: r.age || r.page_age || undefined,
+    extra_snippets: r.extra_snippets || undefined,
+  };
+}
+
+/**
+ * Deduplicate results by URL (first occurrence wins).
+ */
+function deduplicateResults(results: SearchResultFormatted[]): SearchResultFormatted[] {
+  const seen = new Map<string, SearchResultFormatted>();
+  for (const result of results) {
+    const key = toDedupeKey(result.url);
+    if (key !== null && !seen.has(key)) {
+      seen.set(key, result);
+    }
+  }
+  return Array.from(seen.values());
+}
+
+/**
+ * Fetch AI summary from Brave Summarizer API.
+ * This endpoint is free (not billed separately).
+ */
+async function fetchSummary(
+  summarizerKey: string,
+  signal?: AbortSignal
+): Promise<string | null> {
+  const cached = summarizerCache.get(summarizerKey);
+  if (cached !== undefined) return cached;
+
+  try {
+    const url = `https://api.search.brave.com/res/v1/summarizer/search?key=${encodeURIComponent(summarizerKey)}&entity_info=false`;
+    const response = await fetchWithRetry(url, {
+      method: "GET",
+      headers: braveHeaders(),
+      signal,
+    }, 1);
+
+    const data: BraveSummarizerResponse = await response.json();
+
+    // Extract summary text from the structured response
+    let summaryText = "";
+    if (data.summary && Array.isArray(data.summary)) {
+      summaryText = data.summary
+        .filter((s) => s.type === "token" || s.type === "text")
+        .map((s) => s.data)
+        .join("");
+    }
+
+    if (summaryText) {
+      summarizerCache.set(summarizerKey, summaryText);
+      return summaryText;
+    }
+    return null;
+  } catch {
+    // Summarizer is best-effort — don't fail the search
+    return null;
+  }
+}
+
+// =============================================================================
+// Tool Registration
+// =============================================================================
+
+export function registerSearchTool(pi: ExtensionAPI) {
+  pi.registerTool({
+    name: "search-the-web",
+    label: "Web Search",
+    description:
+      "Search the web using Brave Search API. Returns top results with titles, URLs, descriptions, " +
+      "extra contextual snippets, result ages, and optional AI summary. " +
+      "Supports freshness filtering, domain filtering, and auto-detects recency-sensitive queries.",
+    promptSnippet: "Search the web for information",
+    promptGuidelines: [
+      "Use this tool when the user asks about current events, facts, or external knowledge not in the codebase.",
+      "Always provide the search query to the user in your response.",
+      "Limit to 3-5 results unless more context is needed.",
+      "Use freshness='week' or 'month' for queries about recent events, releases, or updates.",
+      "Use the fetch_page tool to read the full content of promising URLs from search results.",
+    ],
+    parameters: Type.Object({
+      query: Type.String({ description: "Search query (e.g., 'latest AI news')" }),
+      count: Type.Optional(
+        Type.Number({ minimum: 1, maximum: 10, default: 5, description: "Number of results to return (default: 5)" })
+      ),
+      freshness: Type.Optional(
+        StringEnum(["auto", "day", "week", "month", "year"] as const, {
+          description:
+            "Filter by recency. 'auto' (default) detects from query. 'day'=past 24h, 'week'=past 7d, 'month'=past 30d, 'year'=past 365d.",
+        })
+      ),
+      domain: Type.Optional(
+        Type.String({
+          description: "Limit results to a specific domain (e.g., 'stackoverflow.com', 'github.com')",
+        })
+      ),
+      summary: Type.Optional(
+        Type.Boolean({
+          description: "Request an AI-generated summary of the search results (default: false). Adds latency but provides a concise answer.",
+          default: false,
+        })
+      ),
+    }),
+
+    async execute(toolCallId, params, signal, onUpdate, ctx) {
+      if (signal?.aborted) {
+        return { content: [{ type: "text", text: "Search cancelled." }] };
+      }
+
+      const apiKey = getBraveApiKey();
+      if (!apiKey) {
+        return {
+          content: [{ type: "text", text: "Web search unavailable: BRAVE_API_KEY is not set. Set the BRAVE_API_KEY environment variable." }],
+          isError: true,
+        };
+      }
+
+      const count = params.count ?? 5;
+      const wantSummary = params.summary ?? false;
+
+      // ------------------------------------------------------------------
+      // Resolve freshness — auto-detect if not explicit
+      // ------------------------------------------------------------------
+      let freshness: string | null = null;
+      if (params.freshness && params.freshness !== "auto") {
+        const freshnessMap: Record<string, string> = {
+          day: "pd",
+          week: "pw",
+          month: "pm",
+          year: "py",
+        };
+        freshness = freshnessMap[params.freshness] || null;
+      } else {
+        // Auto-detect recency from query
+        freshness = detectFreshness(params.query);
+      }
+
+      // ------------------------------------------------------------------
+      // Handle domain filter — prepend site: to query if specified
+      // ------------------------------------------------------------------
+      let effectiveQuery = params.query;
+      if (params.domain) {
+        // Only add site: filter if not already in query
+        if (!effectiveQuery.toLowerCase().includes("site:")) {
+          effectiveQuery = `site:${params.domain} ${effectiveQuery}`;
+        }
+      }
+
+      // ------------------------------------------------------------------
+      // Cache lookup
+      // ------------------------------------------------------------------
+      const cacheKey = normalizeQuery(effectiveQuery) + `|f:${freshness || ""}|s:${wantSummary}`;
+      const cached = searchCache.get(cacheKey);
+
+      if (cached) {
+        const limited = cached.results.slice(0, count);
+
+        // Try to get summary from cache too
+        let summaryText: string | undefined;
+        if (wantSummary && cached.summarizerKey) {
+          summaryText = (await fetchSummary(cached.summarizerKey, signal)) ?? undefined;
+        }
+
+        const output = formatSearchResults(params.query, limited, {
+          cached: true,
+          summary: summaryText,
+        });
+
+        const truncation = truncateHead(output, { maxLines: DEFAULT_MAX_LINES, maxBytes: DEFAULT_MAX_BYTES });
+        let content = truncation.content;
+        if (truncation.truncated) {
+          const tempFile = await pi.writeTempFile(output, { prefix: "web-search-" });
+          content += `\n\n[Truncated: ${truncation.outputLines}/${truncation.totalLines} lines (${formatSize(truncation.outputBytes)}/${formatSize(truncation.totalBytes)}). Full results: ${tempFile}]`;
+        }
+
+        return {
+          content: [{ type: "text", text: content }],
+          details: {
+            query: params.query,
+            effectiveQuery,
+            results: limited,
+            count: limited.length,
+            cached: true,
+            freshness: freshness || "none",
+            hasSummary: !!summaryText,
+          },
+        };
+      }
+
+      onUpdate?.({ content: [{ type: "text", text: `Searching for "${params.query}"...` }] });
+
+      try {
+        // ------------------------------------------------------------------
+        // Build Brave API request
+        // ------------------------------------------------------------------
+        const url = new URL("https://api.search.brave.com/res/v1/web/search");
+        url.searchParams.append("q", effectiveQuery);
+        url.searchParams.append("count", "10"); // Fetch extra for dedup headroom
+        url.searchParams.append("extra_snippets", "true");
+        url.searchParams.append("text_decorations", "false"); // No HTML bold tags in snippets
+
+        if (freshness) {
+          url.searchParams.append("freshness", freshness);
+        }
+
+        if (wantSummary) {
+          url.searchParams.append("summary", "1");
+        }
+
+        const fetchOptions: RequestInit = {
+          method: "GET",
+          headers: braveHeaders(),
+          signal,
+        };
+
+        let response: Response;
+        try {
+          response = await fetchWithRetry(url.toString(), fetchOptions, 2);
+        } catch (fetchErr) {
+          const message = fetchErr instanceof HttpError
+            ? `HTTP ${fetchErr.statusCode}: ${fetchErr.message}`
+            : (fetchErr as Error).message ?? String(fetchErr);
+          return {
+            content: [{ type: "text", text: `Search failed: ${message}` }],
+            details: { error: message, query: params.query },
+            isError: true,
+          };
+        }
+
+        const data = await response.json();
+        const rawResults: BraveWebResult[] = data.web?.results ?? [];
+        const summarizerKey: string | undefined = data.summarizer?.key;
+
+        // ------------------------------------------------------------------
+        // Normalize, deduplicate, cache
+        // ------------------------------------------------------------------
+        const normalized = rawResults.map(normalizeBraveResult);
+        const deduplicated = deduplicateResults(normalized);
+
+        searchCache.set(cacheKey, { results: deduplicated, summarizerKey });
+
+        const results = deduplicated.slice(0, count);
+
+        // ------------------------------------------------------------------
+        // Optionally fetch AI summary (non-blocking — best-effort)
+        // ------------------------------------------------------------------
+        let summaryText: string | undefined;
+        if (wantSummary && summarizerKey) {
+          summaryText = (await fetchSummary(summarizerKey, signal)) ?? undefined;
+        }
+
+        // ------------------------------------------------------------------
+        // Format output
+        // ------------------------------------------------------------------
+        const output = formatSearchResults(params.query, results, {
+          summary: summaryText,
+        });
+
+        const truncation = truncateHead(output, { maxLines: DEFAULT_MAX_LINES, maxBytes: DEFAULT_MAX_BYTES });
+        let content = truncation.content;
+
+        if (truncation.truncated) {
+          const tempFile = await pi.writeTempFile(output, { prefix: "web-search-" });
+          content += `\n\n[Truncated: ${truncation.outputLines}/${truncation.totalLines} lines (${formatSize(truncation.outputBytes)}/${formatSize(truncation.totalBytes)}). Full results: ${tempFile}]`;
+        }
+
+        return {
+          content: [{ type: "text", text: content }],
+          details: {
+            query: params.query,
+            effectiveQuery,
+            results,
+            count: results.length,
+            cached: false,
+            freshness: freshness || "none",
+            hasSummary: !!summaryText,
+          },
+        };
+      } catch (error) {
+        return {
+          content: [{ type: "text", text: `Search failed: ${(error as Error).message}` }],
+          details: { error: (error as Error).message, query: params.query },
+          isError: true,
+        };
+      }
+    },
+
+    renderCall(args, theme) {
+      let text = theme.fg("toolTitle", theme.bold("search-the-web "));
+      text += theme.fg("muted", `"${args.query}"`);
+
+      const meta: string[] = [];
+      if (args.count && args.count !== 5) meta.push(`${args.count} results`);
+      if (args.freshness && args.freshness !== "auto") meta.push(`freshness:${args.freshness}`);
+      if (args.domain) meta.push(`site:${args.domain}`);
+      if (args.summary) meta.push("+ summary");
+      if (meta.length > 0) {
+        text += " " + theme.fg("dim", `(${meta.join(", ")})`);
+      }
+
+      return new Text(text, 0, 0);
+    },
+
+    renderResult(result, { expanded }, theme) {
+      const details = result.details as any;
+      if (details?.error) {
+        return new Text(theme.fg("error", `✗ Search failed: ${details.error}`), 0, 0);
+      }
+
+      const cacheTag = details?.cached ? theme.fg("dim", " [cached]") : "";
+      const freshTag = details?.freshness && details.freshness !== "none"
+        ? theme.fg("dim", ` [${details.freshness}]`)
+        : "";
+      const summaryTag = details?.hasSummary ? theme.fg("dim", " [+summary]") : "";
+
+      let text = theme.fg("success", `✓ ${details?.count ?? 0} results for "${details?.query}"`) +
+        cacheTag + freshTag + summaryTag;
+
+      if (expanded && details?.results) {
+        text += "\n\n";
+        for (const r of details.results.slice(0, 3)) {
+          const age = r.age ? theme.fg("dim", ` (${r.age})`) : "";
+          text += `${theme.bold(r.title)}${age}\n${r.url}\n${r.description}\n\n`;
+        }
+        if (details.results.length > 3) {
+          text += theme.fg("dim", `... and ${details.results.length - 3} more`);
+        }
+      }
+
+      return new Text(text, 0, 0);
+    },
+  });
+}

package/src/resources/extensions/search-the-web/url-utils.ts

@@ -0,0 +1,91 @@
+/**
+ * URL normalization and query utilities.
+ */
+
+/** Normalize a search query into a stable cache key. */
+export function normalizeQuery(query: string): string {
+  return query.trim().toLowerCase().replace(/\s+/g, " ").normalize("NFC");
+}
+
+/**
+ * Canonical URL for deduplication.
+ * Strips fragment, tracking params, lowercases hostname, sorts query params,
+ * strips trailing "/" on root paths.
+ */
+export function toDedupeKey(url: string): string | null {
+  try {
+    const parsed = new URL(url);
+    parsed.hostname = parsed.hostname.toLowerCase();
+    parsed.hash = "";
+
+    const TRACKING_PARAMS = new Set(["fbclid", "gclid"]);
+    const toDelete: string[] = [];
+    for (const key of parsed.searchParams.keys()) {
+      if (key.startsWith("utm_") || TRACKING_PARAMS.has(key)) {
+        toDelete.push(key);
+      }
+    }
+    for (const key of toDelete) parsed.searchParams.delete(key);
+    parsed.searchParams.sort();
+
+    let canonical = parsed.toString();
+    if (parsed.pathname === "/" && !parsed.search) {
+      canonical = canonical.replace(/\/$/, "");
+    }
+    return canonical;
+  } catch {
+    return null;
+  }
+}
+
+/**
+ * Extract a clean domain from a URL for display.
+ * "https://docs.python.org/3/library/asyncio.html" → "docs.python.org"
+ */
+export function extractDomain(url: string): string {
+  try {
+    return new URL(url).hostname.replace(/^www\./, "");
+  } catch {
+    return url;
+  }
+}
+
+/**
+ * Detect if a query likely wants fresh/recent results.
+ * Returns a suggested Brave freshness parameter or null.
+ */
+export function detectFreshness(query: string): string | null {
+  const q = query.toLowerCase();
+
+  // Explicit year references for current/recent years
+  const currentYear = new Date().getFullYear();
+  for (let y = currentYear; y >= currentYear - 1; y--) {
+    if (q.includes(String(y))) return "py"; // past year
+  }
+
+  // Recency keywords
+  const recentPatterns = [
+    /\b(latest|newest|recent|new|just released|just launched)\b/,
+    /\b(today|yesterday|this week|this month)\b/,
+    /\b(breaking|update|announcement|release notes?)\b/,
+    /\b(what('?s| is) new)\b/,
+  ];
+  for (const pattern of recentPatterns) {
+    if (pattern.test(q)) return "pm"; // past month
+  }
+
+  return null;
+}
+
+/**
+ * Detect if a query targets specific domains.
+ * Returns extracted domains or null.
+ */
+export function detectDomainHints(query: string): string[] | null {
+  // Match "site:example.com" patterns
+  const siteMatches = query.match(/site:(\S+)/gi);
+  if (siteMatches) {
+    return siteMatches.map((m) => m.replace(/^site:/i, ""));
+  }
+  return null;
+}