@askalf/deepdive 0.1.0

package/dist/search/brave.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"brave.js","sourceRoot":"","sources":["../../src/search/brave.ts"],"names":[],"mappings":"AAAA,yDAAyD;AAIzD,MAAM,OAAO,WAAW;IAEO;IADpB,IAAI,GAAG,OAAO,CAAC;IACxB,YAA6B,GAAW;QAAX,QAAG,GAAH,GAAG,CAAQ;IAAG,CAAC;IAE5C,KAAK,CAAC,MAAM,CAAC,KAAa,EAAE,KAAa,EAAE,MAAoB;QAC7D,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,gDAAgD,CAAC,CAAC;QACtE,GAAG,CAAC,YAAY,CAAC,GAAG,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;QACjC,GAAG,CAAC,YAAY,CAAC,GAAG,CAAC,OAAO,EAAE,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;QAC3D,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,GAAG,EAAE;YAC3B,OAAO,EAAE;gBACP,MAAM,EAAE,kBAAkB;gBAC1B,sBAAsB,EAAE,IAAI,CAAC,GAAG;aACjC;YACD,MAAM;SACP,CAAC,CAAC;QACH,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,IAAI,KAAK,CAAC,SAAS,GAAG,CAAC,MAAM,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC,CAAC;QACtE,MAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAE7B,CAAC;QACF,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,EAAE,OAAO,IAAI,EAAE,CAAC;QACtC,OAAO,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;YAC1B,GAAG,EAAE,CAAC,CAAC,GAAG;YACV,KAAK,EAAE,CAAC,CAAC,KAAK,IAAI,EAAE;YACpB,OAAO,EAAE,CAAC,CAAC,WAAW,IAAI,EAAE;YAC5B,IAAI,EAAE,CAAC,GAAG,CAAC;SACZ,CAAC,CAAC,CAAC;IACN,CAAC;CACF"}

package/dist/search/duckduckgo.d.ts

@@ -0,0 +1,7 @@
+import type { SearchAdapter, SearchResult } from "../search.js";
+export declare class DuckDuckGoSearch implements SearchAdapter {
+    readonly name = "duckduckgo";
+    search(query: string, limit: number, signal?: AbortSignal): Promise<SearchResult[]>;
+}
+export declare function parseDuckDuckGoHTML(html: string, limit: number): SearchResult[];
+//# sourceMappingURL=duckduckgo.d.ts.map

package/dist/search/duckduckgo.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"duckduckgo.d.ts","sourceRoot":"","sources":["../../src/search/duckduckgo.ts"],"names":[],"mappings":"AAOA,OAAO,KAAK,EAAE,aAAa,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAIhE,qBAAa,gBAAiB,YAAW,aAAa;IACpD,QAAQ,CAAC,IAAI,gBAAgB;IAEvB,MAAM,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,YAAY,EAAE,CAAC;CAiB1F;AAGD,wBAAgB,mBAAmB,CAAC,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,GAAG,YAAY,EAAE,CAuB/E"}

package/dist/search/duckduckgo.js

@@ -0,0 +1,94 @@
+// DuckDuckGo HTML search. Uses the html.duckduckgo.com endpoint, which returns
+// a static HTML document that's scrapable without an API key. Parses result
+// links from the document structure, unwraps DDG's //duckduckgo.com/l/?uddg=
+// redirect tracking, and returns { url, title, snippet }.
+//
+// If DDG changes their HTML layout, this breaks. Regenerate parser accordingly.
+const ENDPOINT = "https://html.duckduckgo.com/html/";
+export class DuckDuckGoSearch {
+    name = "duckduckgo";
+    async search(query, limit, signal) {
+        const form = new URLSearchParams({ q: query });
+        const res = await fetch(ENDPOINT, {
+            method: "POST",
+            headers: {
+                "content-type": "application/x-www-form-urlencoded",
+                "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 " +
+                    "(KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
+            },
+            body: form.toString(),
+            signal,
+        });
+        if (!res.ok)
+            throw new Error(`duckduckgo ${res.status} ${res.statusText}`);
+        const html = await res.text();
+        return parseDuckDuckGoHTML(html, limit);
+    }
+}
+// Exported for unit tests.
+export function parseDuckDuckGoHTML(html, limit) {
+    const results = [];
+    // Each result is wrapped in <div class="result results_links ...">.
+    // We match non-greedily on the title/snippet anchors within.
+    const blockRe = /<div class="result[^"]*"[\s\S]*?<a[^>]+class="result__a"[^>]+href="([^"]+)"[^>]*>([\s\S]*?)<\/a>[\s\S]*?<a[^>]+class="result__snippet"[^>]*>([\s\S]*?)<\/a>/g;
+    let rank = 0;
+    let m;
+    while ((m = blockRe.exec(html)) !== null) {
+        const [, rawHref, rawTitle, rawSnippet] = m;
+        const url = unwrapDDGRedirect(decodeHtmlEntities(rawHref));
+        if (!isValidHttpUrl(url))
+            continue;
+        rank++;
+        results.push({
+            url,
+            title: stripTags(decodeHtmlEntities(rawTitle)).trim(),
+            snippet: stripTags(decodeHtmlEntities(rawSnippet)).trim(),
+            rank,
+        });
+        if (results.length >= limit)
+            break;
+    }
+    return results;
+}
+function unwrapDDGRedirect(href) {
+    // DDG wraps real URLs behind //duckduckgo.com/l/?uddg=<encoded>&rut=<hash>
+    // The real target is in the uddg param.
+    if (href.startsWith("//"))
+        href = "https:" + href;
+    try {
+        const u = new URL(href);
+        if (u.hostname.endsWith("duckduckgo.com") && u.pathname === "/l/") {
+            const uddg = u.searchParams.get("uddg");
+            if (uddg)
+                return decodeURIComponent(uddg);
+        }
+        return u.toString();
+    }
+    catch {
+        return href;
+    }
+}
+function isValidHttpUrl(s) {
+    try {
+        const u = new URL(s);
+        return u.protocol === "http:" || u.protocol === "https:";
+    }
+    catch {
+        return false;
+    }
+}
+function decodeHtmlEntities(s) {
+    return s
+        .replace(/&amp;/g, "&")
+        .replace(/&lt;/g, "<")
+        .replace(/&gt;/g, ">")
+        .replace(/&quot;/g, '"')
+        .replace(/&#39;/g, "'")
+        .replace(/&nbsp;/g, " ")
+        .replace(/&#x2F;/g, "/")
+        .replace(/&#(\d+);/g, (_, n) => String.fromCharCode(parseInt(n, 10)));
+}
+function stripTags(s) {
+    return s.replace(/<[^>]+>/g, "");
+}
+//# sourceMappingURL=duckduckgo.js.map

package/dist/search/duckduckgo.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"duckduckgo.js","sourceRoot":"","sources":["../../src/search/duckduckgo.ts"],"names":[],"mappings":"AAAA,+EAA+E;AAC/E,4EAA4E;AAC5E,6EAA6E;AAC7E,0DAA0D;AAC1D,EAAE;AACF,gFAAgF;AAIhF,MAAM,QAAQ,GAAG,mCAAmC,CAAC;AAErD,MAAM,OAAO,gBAAgB;IAClB,IAAI,GAAG,YAAY,CAAC;IAE7B,KAAK,CAAC,MAAM,CAAC,KAAa,EAAE,KAAa,EAAE,MAAoB;QAC7D,MAAM,IAAI,GAAG,IAAI,eAAe,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC;QAC/C,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,QAAQ,EAAE;YAChC,MAAM,EAAE,MAAM;YACd,OAAO,EAAE;gBACP,cAAc,EAAE,mCAAmC;gBACnD,YAAY,EACV,+DAA+D;oBAC/D,oDAAoD;aACvD;YACD,IAAI,EAAE,IAAI,CAAC,QAAQ,EAAE;YACrB,MAAM;SACP,CAAC,CAAC;QACH,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,IAAI,KAAK,CAAC,cAAc,GAAG,CAAC,MAAM,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC,CAAC;QAC3E,MAAM,IAAI,GAAG,MAAM,GAAG,CAAC,IAAI,EAAE,CAAC;QAC9B,OAAO,mBAAmB,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC;IAC1C,CAAC;CACF;AAED,2BAA2B;AAC3B,MAAM,UAAU,mBAAmB,CAAC,IAAY,EAAE,KAAa;IAC7D,MAAM,OAAO,GAAmB,EAAE,CAAC;IACnC,oEAAoE;IACpE,6DAA6D;IAC7D,MAAM,OAAO,GACX,8JAA8J,CAAC;IAEjK,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,IAAI,CAAyB,CAAC;IAC9B,OAAO,CAAC,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACzC,MAAM,CAAC,EAAE,OAAO,EAAE,QAAQ,EAAE,UAAU,CAAC,GAAG,CAAC,CAAC;QAC5C,MAAM,GAAG,GAAG,iBAAiB,CAAC,kBAAkB,CAAC,OAAO,CAAC,CAAC,CAAC;QAC3D,IAAI,CAAC,cAAc,CAAC,GAAG,CAAC;YAAE,SAAS;QACnC,IAAI,EAAE,CAAC;QACP,OAAO,CAAC,IAAI,CAAC;YACX,GAAG;YACH,KAAK,EAAE,SAAS,CAAC,kBAAkB,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,EAAE;YACrD,OAAO,EAAE,SAAS,CAAC,kBAAkB,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,EAAE;YACzD,IAAI;SACL,CAAC,CAAC;QACH,IAAI,OAAO,CAAC,MAAM,IAAI,KAAK;YAAE,MAAM;IACrC,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,SAAS,iBAAiB,CAAC,IAAY;IACrC,2EAA2E;IAC3E,wCAAwC;IACxC,IAAI,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC;QAAE,IAAI,GAAG,QAAQ,GAAG,IAAI,CAAC;IAClD,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,IAAI,CAAC,CAAC;QACxB,IAAI,CAAC,CAAC,QAAQ,CAAC,QAAQ,CAAC,gBAAgB,CAAC,IAAI,CAAC,CAAC,QAAQ,KAAK,KAAK,EAAE,CAAC;YAClE,MAAM,IAAI,GAAG,CAAC,CAAC,YAAY,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;YACxC,IAAI,IAAI;gBAAE,OAAO,kBAAkB,CAAC,IAAI,CAAC,CAAC;QAC5C,CAAC;QACD,OAAO,CAAC,CAAC,QAAQ,EAAE,CAAC;IACtB,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED,SAAS,cAAc,CAAC,CAAS;IAC/B,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC;QACrB,OAAO,CAAC,CAAC,QAAQ,KAAK,OAAO,IAAI,CAAC,CAAC,QAAQ,KAAK,QAAQ,CAAC;IAC3D,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED,SAAS,kBAAkB,CAAC,CAAS;IACnC,OAAO,CAAC;SACL,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC;SACtB,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC;SACtB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,WAAW,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,MAAM,CAAC,YAAY,CAAC,QAAQ,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC;AAC1E,CAAC;AAED,SAAS,SAAS,CAAC,CAAS;IAC1B,OAAO,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC;AACnC,CAAC"}

package/dist/search/searxng.d.ts

@@ -0,0 +1,8 @@
+import type { SearchAdapter, SearchResult } from "../search.js";
+export declare class SearXNGSearch implements SearchAdapter {
+    private readonly baseUrl;
+    readonly name = "searxng";
+    constructor(baseUrl: string);
+    search(query: string, limit: number, signal?: AbortSignal): Promise<SearchResult[]>;
+}
+//# sourceMappingURL=searxng.d.ts.map

package/dist/search/searxng.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"searxng.d.ts","sourceRoot":"","sources":["../../src/search/searxng.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,aAAa,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAEhE,qBAAa,aAAc,YAAW,aAAa;IAErC,OAAO,CAAC,QAAQ,CAAC,OAAO;IADpC,QAAQ,CAAC,IAAI,aAAa;gBACG,OAAO,EAAE,MAAM;IAEtC,MAAM,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,YAAY,EAAE,CAAC;CAoB1F"}

package/dist/search/searxng.js

@@ -0,0 +1,29 @@
+// SearXNG adapter. Points at an existing SearXNG instance (self-hosted or
+// public). Requires DEEPDIVE_SEARXNG_URL. Uses the JSON output format.
+export class SearXNGSearch {
+    baseUrl;
+    name = "searxng";
+    constructor(baseUrl) {
+        this.baseUrl = baseUrl;
+    }
+    async search(query, limit, signal) {
+        const url = new URL(this.baseUrl.replace(/\/+$/, "") + "/search");
+        url.searchParams.set("q", query);
+        url.searchParams.set("format", "json");
+        const res = await fetch(url, {
+            headers: { accept: "application/json" },
+            signal,
+        });
+        if (!res.ok)
+            throw new Error(`searxng ${res.status} ${res.statusText}`);
+        const json = (await res.json());
+        const items = json.results ?? [];
+        return items.slice(0, limit).map((r, i) => ({
+            url: r.url,
+            title: r.title ?? "",
+            snippet: r.content ?? "",
+            rank: i + 1,
+        }));
+    }
+}
+//# sourceMappingURL=searxng.js.map

package/dist/search/searxng.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"searxng.js","sourceRoot":"","sources":["../../src/search/searxng.ts"],"names":[],"mappings":"AAAA,0EAA0E;AAC1E,uEAAuE;AAIvE,MAAM,OAAO,aAAa;IAEK;IADpB,IAAI,GAAG,SAAS,CAAC;IAC1B,YAA6B,OAAe;QAAf,YAAO,GAAP,OAAO,CAAQ;IAAG,CAAC;IAEhD,KAAK,CAAC,MAAM,CAAC,KAAa,EAAE,KAAa,EAAE,MAAoB;QAC7D,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,GAAG,SAAS,CAAC,CAAC;QAClE,GAAG,CAAC,YAAY,CAAC,GAAG,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;QACjC,GAAG,CAAC,YAAY,CAAC,GAAG,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;QACvC,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,GAAG,EAAE;YAC3B,OAAO,EAAE,EAAE,MAAM,EAAE,kBAAkB,EAAE;YACvC,MAAM;SACP,CAAC,CAAC;QACH,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,IAAI,KAAK,CAAC,WAAW,GAAG,CAAC,MAAM,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC,CAAC;QACxE,MAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAE7B,CAAC;QACF,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,IAAI,EAAE,CAAC;QACjC,OAAO,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;YAC1C,GAAG,EAAE,CAAC,CAAC,GAAG;YACV,KAAK,EAAE,CAAC,CAAC,KAAK,IAAI,EAAE;YACpB,OAAO,EAAE,CAAC,CAAC,OAAO,IAAI,EAAE;YACxB,IAAI,EAAE,CAAC,GAAG,CAAC;SACZ,CAAC,CAAC,CAAC;IACN,CAAC;CACF"}

package/dist/search/tavily.d.ts

@@ -0,0 +1,8 @@
+import type { SearchAdapter, SearchResult } from "../search.js";
+export declare class TavilySearch implements SearchAdapter {
+    private readonly key;
+    readonly name = "tavily";
+    constructor(key: string);
+    search(query: string, limit: number, signal?: AbortSignal): Promise<SearchResult[]>;
+}
+//# sourceMappingURL=tavily.d.ts.map

package/dist/search/tavily.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"tavily.d.ts","sourceRoot":"","sources":["../../src/search/tavily.ts"],"names":[],"mappings":"AAKA,OAAO,KAAK,EAAE,aAAa,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAEhE,qBAAa,YAAa,YAAW,aAAa;IAEpC,OAAO,CAAC,QAAQ,CAAC,GAAG;IADhC,QAAQ,CAAC,IAAI,YAAY;gBACI,GAAG,EAAE,MAAM;IAElC,MAAM,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,YAAY,EAAE,CAAC;CA2B1F"}

package/dist/search/tavily.js

@@ -0,0 +1,38 @@
+// Tavily adapter. Requires DEEPDIVE_TAVILY_KEY. Tavily returns search results
+// with pre-extracted content, which means the browser fetch step can be
+// skipped when using this adapter — but deepdive re-fetches anyway so the
+// downstream extract/synthesis path is identical across adapters.
+export class TavilySearch {
+    key;
+    name = "tavily";
+    constructor(key) {
+        this.key = key;
+    }
+    async search(query, limit, signal) {
+        const res = await fetch("https://api.tavily.com/search", {
+            method: "POST",
+            headers: {
+                "content-type": "application/json",
+                accept: "application/json",
+            },
+            body: JSON.stringify({
+                api_key: this.key,
+                query,
+                max_results: limit,
+                search_depth: "basic",
+            }),
+            signal,
+        });
+        if (!res.ok)
+            throw new Error(`tavily ${res.status} ${res.statusText}`);
+        const json = (await res.json());
+        const items = json.results ?? [];
+        return items.slice(0, limit).map((r, i) => ({
+            url: r.url,
+            title: r.title ?? "",
+            snippet: r.content ?? "",
+            rank: i + 1,
+        }));
+    }
+}
+//# sourceMappingURL=tavily.js.map

package/dist/search/tavily.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"tavily.js","sourceRoot":"","sources":["../../src/search/tavily.ts"],"names":[],"mappings":"AAAA,8EAA8E;AAC9E,wEAAwE;AACxE,0EAA0E;AAC1E,kEAAkE;AAIlE,MAAM,OAAO,YAAY;IAEM;IADpB,IAAI,GAAG,QAAQ,CAAC;IACzB,YAA6B,GAAW;QAAX,QAAG,GAAH,GAAG,CAAQ;IAAG,CAAC;IAE5C,KAAK,CAAC,MAAM,CAAC,KAAa,EAAE,KAAa,EAAE,MAAoB;QAC7D,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,+BAA+B,EAAE;YACvD,MAAM,EAAE,MAAM;YACd,OAAO,EAAE;gBACP,cAAc,EAAE,kBAAkB;gBAClC,MAAM,EAAE,kBAAkB;aAC3B;YACD,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC;gBACnB,OAAO,EAAE,IAAI,CAAC,GAAG;gBACjB,KAAK;gBACL,WAAW,EAAE,KAAK;gBAClB,YAAY,EAAE,OAAO;aACtB,CAAC;YACF,MAAM;SACP,CAAC,CAAC;QACH,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,IAAI,KAAK,CAAC,UAAU,GAAG,CAAC,MAAM,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC,CAAC;QACvE,MAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAE7B,CAAC;QACF,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,IAAI,EAAE,CAAC;QACjC,OAAO,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;YAC1C,GAAG,EAAE,CAAC,CAAC,GAAG;YACV,KAAK,EAAE,CAAC,CAAC,KAAK,IAAI,EAAE;YACpB,OAAO,EAAE,CAAC,CAAC,OAAO,IAAI,EAAE;YACxB,IAAI,EAAE,CAAC,GAAG,CAAC;SACZ,CAAC,CAAC,CAAC;IACN,CAAC;CACF"}

package/dist/search.d.ts

@@ -0,0 +1,13 @@
+export interface SearchResult {
+    url: string;
+    title: string;
+    snippet: string;
+    rank: number;
+}
+export interface SearchAdapter {
+    readonly name: string;
+    search(query: string, limit: number, signal?: AbortSignal): Promise<SearchResult[]>;
+}
+export declare function resolveSearchAdapter(name: string, env: Record<string, string | undefined>): Promise<SearchAdapter>;
+export declare function dedupeByUrl(results: SearchResult[]): SearchResult[];
+//# sourceMappingURL=search.d.ts.map

package/dist/search.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"search.d.ts","sourceRoot":"","sources":["../src/search.ts"],"names":[],"mappings":"AAGA,MAAM,WAAW,YAAY;IAC3B,GAAG,EAAE,MAAM,CAAC;IACZ,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC;IAChB,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,aAAa;IAC5B,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,MAAM,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,WAAW,GAAG,OAAO,CAAC,YAAY,EAAE,CAAC,CAAC;CACrF;AAED,wBAAsB,oBAAoB,CACxC,IAAI,EAAE,MAAM,EACZ,GAAG,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC,GACtC,OAAO,CAAC,aAAa,CAAC,CA4BxB;AAED,wBAAgB,WAAW,CAAC,OAAO,EAAE,YAAY,EAAE,GAAG,YAAY,EAAE,CAUnE"}

package/dist/search.js

@@ -0,0 +1,47 @@
+// Search adapter interface. A "search" returns candidate URLs with metadata.
+// Adapters live under src/search/*. Default is DuckDuckGo HTML (no API key).
+export async function resolveSearchAdapter(name, env) {
+    switch (name) {
+        case "duckduckgo":
+        case "ddg": {
+            const { DuckDuckGoSearch } = await import("./search/duckduckgo.js");
+            return new DuckDuckGoSearch();
+        }
+        case "searxng": {
+            const { SearXNGSearch } = await import("./search/searxng.js");
+            const url = env.DEEPDIVE_SEARXNG_URL;
+            if (!url)
+                throw new Error("searxng adapter requires DEEPDIVE_SEARXNG_URL");
+            return new SearXNGSearch(url);
+        }
+        case "brave": {
+            const { BraveSearch } = await import("./search/brave.js");
+            const key = env.DEEPDIVE_BRAVE_KEY;
+            if (!key)
+                throw new Error("brave adapter requires DEEPDIVE_BRAVE_KEY");
+            return new BraveSearch(key);
+        }
+        case "tavily": {
+            const { TavilySearch } = await import("./search/tavily.js");
+            const key = env.DEEPDIVE_TAVILY_KEY;
+            if (!key)
+                throw new Error("tavily adapter requires DEEPDIVE_TAVILY_KEY");
+            return new TavilySearch(key);
+        }
+        default:
+            throw new Error(`unknown search adapter: ${name}`);
+    }
+}
+export function dedupeByUrl(results) {
+    const seen = new Set();
+    const out = [];
+    for (const r of results) {
+        const key = r.url.replace(/#.*$/, "").replace(/\/+$/, "");
+        if (seen.has(key))
+            continue;
+        seen.add(key);
+        out.push(r);
+    }
+    return out;
+}
+//# sourceMappingURL=search.js.map

package/dist/search.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"search.js","sourceRoot":"","sources":["../src/search.ts"],"names":[],"mappings":"AAAA,6EAA6E;AAC7E,6EAA6E;AAc7E,MAAM,CAAC,KAAK,UAAU,oBAAoB,CACxC,IAAY,EACZ,GAAuC;IAEvC,QAAQ,IAAI,EAAE,CAAC;QACb,KAAK,YAAY,CAAC;QAClB,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,gBAAgB,EAAE,GAAG,MAAM,MAAM,CAAC,wBAAwB,CAAC,CAAC;YACpE,OAAO,IAAI,gBAAgB,EAAE,CAAC;QAChC,CAAC;QACD,KAAK,SAAS,CAAC,CAAC,CAAC;YACf,MAAM,EAAE,aAAa,EAAE,GAAG,MAAM,MAAM,CAAC,qBAAqB,CAAC,CAAC;YAC9D,MAAM,GAAG,GAAG,GAAG,CAAC,oBAAoB,CAAC;YACrC,IAAI,CAAC,GAAG;gBAAE,MAAM,IAAI,KAAK,CAAC,+CAA+C,CAAC,CAAC;YAC3E,OAAO,IAAI,aAAa,CAAC,GAAG,CAAC,CAAC;QAChC,CAAC;QACD,KAAK,OAAO,CAAC,CAAC,CAAC;YACb,MAAM,EAAE,WAAW,EAAE,GAAG,MAAM,MAAM,CAAC,mBAAmB,CAAC,CAAC;YAC1D,MAAM,GAAG,GAAG,GAAG,CAAC,kBAAkB,CAAC;YACnC,IAAI,CAAC,GAAG;gBAAE,MAAM,IAAI,KAAK,CAAC,2CAA2C,CAAC,CAAC;YACvE,OAAO,IAAI,WAAW,CAAC,GAAG,CAAC,CAAC;QAC9B,CAAC;QACD,KAAK,QAAQ,CAAC,CAAC,CAAC;YACd,MAAM,EAAE,YAAY,EAAE,GAAG,MAAM,MAAM,CAAC,oBAAoB,CAAC,CAAC;YAC5D,MAAM,GAAG,GAAG,GAAG,CAAC,mBAAmB,CAAC;YACpC,IAAI,CAAC,GAAG;gBAAE,MAAM,IAAI,KAAK,CAAC,6CAA6C,CAAC,CAAC;YACzE,OAAO,IAAI,YAAY,CAAC,GAAG,CAAC,CAAC;QAC/B,CAAC;QACD;YACE,MAAM,IAAI,KAAK,CAAC,2BAA2B,IAAI,EAAE,CAAC,CAAC;IACvD,CAAC;AACH,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,OAAuB;IACjD,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;IAC/B,MAAM,GAAG,GAAmB,EAAE,CAAC;IAC/B,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,GAAG,GAAG,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;QAC1D,IAAI,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;YAAE,SAAS;QAC5B,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QACd,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACd,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC"}

package/dist/synthesize.d.ts

@@ -0,0 +1,8 @@
+import { type LLMConfig } from "./llm.js";
+import type { Source } from "./citations.js";
+export interface SourceWithContent extends Source {
+    content: string;
+}
+export declare function synthesize(question: string, sources: SourceWithContent[], config: LLMConfig, signal?: AbortSignal): Promise<string>;
+export declare function buildSourcePacket(sources: SourceWithContent[]): string;
+//# sourceMappingURL=synthesize.d.ts.map

package/dist/synthesize.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"synthesize.d.ts","sourceRoot":"","sources":["../src/synthesize.ts"],"names":[],"mappings":"AAIA,OAAO,EAAW,KAAK,SAAS,EAAE,MAAM,UAAU,CAAC;AACnD,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,gBAAgB,CAAC;AAE7C,MAAM,WAAW,iBAAkB,SAAQ,MAAM;IAC/C,OAAO,EAAE,MAAM,CAAC;CACjB;AAkBD,wBAAsB,UAAU,CAC9B,QAAQ,EAAE,MAAM,EAChB,OAAO,EAAE,iBAAiB,EAAE,EAC5B,MAAM,EAAE,SAAS,EACjB,MAAM,CAAC,EAAE,WAAW,GACnB,OAAO,CAAC,MAAM,CAAC,CAgBjB;AAGD,wBAAgB,iBAAiB,CAAC,OAAO,EAAE,iBAAiB,EAAE,GAAG,MAAM,CAOtE"}

package/dist/synthesize.js

@@ -0,0 +1,40 @@
+// Final answer synthesis — takes the original question and the collected
+// sources, asks the LLM to produce a cited markdown answer. Sources are
+// passed as a numbered list so the model can cite them inline as [1], [2].
+import { callLLM } from "./llm.js";
+const SYNTH_SYSTEM = `You are a careful research assistant. You will be given:
+1. The user's original question.
+2. A numbered list of source documents with titles, URLs, and extracted text.
+
+Your job: write a direct, factual, well-structured markdown answer to the question, citing sources inline as [N] where N matches the numbered source list.
+
+Rules:
+- Cite every non-trivial claim inline with [N] (multiple sources: [1][3]).
+- Do not invent a source number that is not in the provided list.
+- Do not cite general knowledge; cite the sources.
+- Prefer concrete facts, dates, numbers, names, mechanisms over hedged prose.
+- If the sources disagree, surface the disagreement.
+- If the sources do not answer the question, say so — do not hallucinate.
+- Do not include a "Sources" section yourself — the caller appends it.
+- Length: match the complexity of the question. A one-line question can get a paragraph; a comparison question may need headers and a table.`;
+export async function synthesize(question, sources, config, signal) {
+    if (sources.length === 0) {
+        return "_No sources could be fetched or extracted. Unable to answer._";
+    }
+    const packet = buildSourcePacket(sources);
+    const userMessage = `Question: ${question}\n\n` +
+        `Sources (${sources.length}):\n\n${packet}\n\n` +
+        `Write the cited markdown answer now.`;
+    const { text } = await callLLM([{ role: "user", content: userMessage }], SYNTH_SYSTEM, config, signal);
+    return text;
+}
+// Exported for unit tests.
+export function buildSourcePacket(sources) {
+    return sources
+        .map((s) => {
+        const header = `[${s.id}] ${s.title || "(untitled)"} — ${s.url}`;
+        return `${header}\n\n${s.content}`;
+    })
+        .join("\n\n---\n\n");
+}
+//# sourceMappingURL=synthesize.js.map

package/dist/synthesize.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"synthesize.js","sourceRoot":"","sources":["../src/synthesize.ts"],"names":[],"mappings":"AAAA,yEAAyE;AACzE,wEAAwE;AACxE,2EAA2E;AAE3E,OAAO,EAAE,OAAO,EAAkB,MAAM,UAAU,CAAC;AAOnD,MAAM,YAAY,GAAG;;;;;;;;;;;;;;6IAcwH,CAAC;AAE9I,MAAM,CAAC,KAAK,UAAU,UAAU,CAC9B,QAAgB,EAChB,OAA4B,EAC5B,MAAiB,EACjB,MAAoB;IAEpB,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACzB,OAAO,+DAA+D,CAAC;IACzE,CAAC;IACD,MAAM,MAAM,GAAG,iBAAiB,CAAC,OAAO,CAAC,CAAC;IAC1C,MAAM,WAAW,GACf,aAAa,QAAQ,MAAM;QAC3B,YAAY,OAAO,CAAC,MAAM,SAAS,MAAM,MAAM;QAC/C,sCAAsC,CAAC;IACzC,MAAM,EAAE,IAAI,EAAE,GAAG,MAAM,OAAO,CAC5B,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,WAAW,EAAE,CAAC,EACxC,YAAY,EACZ,MAAM,EACN,MAAM,CACP,CAAC;IACF,OAAO,IAAI,CAAC;AACd,CAAC;AAED,2BAA2B;AAC3B,MAAM,UAAU,iBAAiB,CAAC,OAA4B;IAC5D,OAAO,OAAO;SACX,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE;QACT,MAAM,MAAM,GAAG,IAAI,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,KAAK,IAAI,YAAY,MAAM,CAAC,CAAC,GAAG,EAAE,CAAC;QACjE,OAAO,GAAG,MAAM,OAAO,CAAC,CAAC,OAAO,EAAE,CAAC;IACrC,CAAC,CAAC;SACD,IAAI,CAAC,aAAa,CAAC,CAAC;AACzB,CAAC"}

package/package.json

@@ -0,0 +1,67 @@
+{
+  "name": "@askalf/deepdive",
+  "version": "0.1.0",
+  "description": "A local research agent. One command, cited answer. Routes every LLM call through your own proxy (dario, Anthropic-compat, OpenAI-compat). Headless browser + pluggable search + multi-provider LLM — zero hosted dependencies.",
+  "type": "module",
+  "bin": {
+    "deepdive": "./dist/cli.js"
+  },
+  "main": "./dist/index.js",
+  "types": "./dist/index.d.ts",
+  "exports": {
+    ".": {
+      "import": "./dist/index.js",
+      "types": "./dist/index.d.ts"
+    }
+  },
+  "files": [
+    "dist",
+    "README.md",
+    "LICENSE",
+    "DISCLAIMER.md"
+  ],
+  "scripts": {
+    "build": "tsc",
+    "test": "node --test --test-concurrency=4 test/*.test.mjs",
+    "dev": "tsx src/cli.ts",
+    "start": "node dist/cli.js",
+    "prepublishOnly": "npm run build"
+  },
+  "keywords": [
+    "research",
+    "agent",
+    "llm",
+    "local",
+    "cli",
+    "anthropic",
+    "openai",
+    "claude",
+    "search",
+    "deep-research",
+    "perplexity-alternative",
+    "cite",
+    "citations",
+    "self-hosted"
+  ],
+  "author": "askalf (https://github.com/askalf)",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/askalf/deepdive.git"
+  },
+  "homepage": "https://github.com/askalf/deepdive",
+  "bugs": {
+    "url": "https://github.com/askalf/deepdive/issues"
+  },
+  "engines": {
+    "node": ">=20.0.0"
+  },
+  "dependencies": {
+    "playwright": "^1.47.0"
+  },
+  "devDependencies": {
+    "@types/node": "^22.0.0",
+    "tsx": "^4.19.0",
+    "typescript": "^5.7.0"
+  }
+}