gsd-pi 2.33.1 → 2.34.0-dev.6769913

package/dist/resources/extensions/google-search/index.js

@@ -0,0 +1,373 @@
+/**
+ * Google Search Extension
+ *
+ * Provides a `google_search` tool that performs web searches via Gemini's
+ * Google Search grounding feature. Uses the user's existing GEMINI_API_KEY
+ * and Google Cloud GenAI credits.
+ *
+ * The tool sends queries to Gemini Flash with `googleSearch: {}` enabled.
+ * Gemini internally performs Google searches, synthesizes an answer, and
+ * returns it with source URLs from grounding metadata.
+ */
+import { DEFAULT_MAX_BYTES, DEFAULT_MAX_LINES, formatSize, truncateHead, } from "@gsd/pi-coding-agent";
+import { Text } from "@gsd/pi-tui";
+import { Type } from "@sinclair/typebox";
+let client = null;
+async function getClient() {
+    if (!client) {
+        const { GoogleGenAI } = await import("@google/genai");
+        client = new GoogleGenAI({ apiKey: process.env.GEMINI_API_KEY });
+    }
+    return client;
+}
+/**
+ * Perform a search using OAuth credentials via the Cloud Code Assist API.
+ * This is used as a fallback when GEMINI_API_KEY is not set.
+ */
+async function searchWithOAuth(query, accessToken, projectId, signal) {
+    const model = process.env.GEMINI_SEARCH_MODEL || "gemini-2.5-flash";
+    const url = `https://cloudcode-pa.googleapis.com/v1internal:streamGenerateContent`;
+    const GEMINI_CLI_HEADERS = {
+        ideType: "IDE_UNSPECIFIED",
+        platform: "PLATFORM_UNSPECIFIED",
+        pluginType: "GEMINI",
+    };
+    const executeFetch = async (retries = 3) => {
+        const response = await fetch(url, {
+            method: "POST",
+            headers: {
+                Authorization: `Bearer ${accessToken}`,
+                "Content-Type": "application/json",
+                "User-Agent": "google-cloud-sdk vscode_cloudshelleditor/0.1",
+                "X-Goog-Api-Client": "gl-node/22.17.0",
+                "Client-Metadata": JSON.stringify(GEMINI_CLI_HEADERS),
+            },
+            body: JSON.stringify({
+                project: projectId,
+                model,
+                request: {
+                    contents: [{ parts: [{ text: query }] }],
+                    tools: [{ googleSearch: {} }],
+                },
+            }),
+            signal,
+        });
+        if (!response.ok && retries > 0 && (response.status === 429 || response.status >= 500)) {
+            await new Promise((resolve) => setTimeout(resolve, 1000 * (4 - retries)));
+            return executeFetch(retries - 1);
+        }
+        return response;
+    };
+    const response = await executeFetch();
+    if (!response.ok) {
+        const errorText = await response.text();
+        throw new Error(`Cloud Code Assist API error (${response.status}): ${errorText}`);
+    }
+    // Note: streamGenerateContent returns SSE; for now, we consume all chunks.
+    // For simplicity and to match the previous structure, we'll read to end.
+    const text = await response.text();
+    const jsonLines = text.split("\n")
+        .filter(l => l.startsWith("data:"))
+        .map(l => l.slice(5).trim())
+        .filter(l => l.length > 0);
+    let data;
+    if (jsonLines.length > 0) {
+        // Aggregate chunks if needed, but for now we take the last chunk or assume it's one
+        data = JSON.parse(jsonLines[jsonLines.length - 1]);
+    }
+    else {
+        data = JSON.parse(text);
+    }
+    const candidate = data.response?.candidates?.[0];
+    const answer = candidate?.content?.parts?.find((p) => p.text)?.text ?? "";
+    const grounding = candidate?.groundingMetadata;
+    const sources = [];
+    const seenTitles = new Set();
+    if (grounding?.groundingChunks) {
+        for (const chunk of grounding.groundingChunks) {
+            if (chunk.web) {
+                const title = chunk.web.title ?? "Untitled";
+                if (seenTitles.has(title))
+                    continue;
+                seenTitles.add(title);
+                const domain = chunk.web.domain ?? title;
+                sources.push({
+                    title,
+                    uri: chunk.web.uri ?? "",
+                    domain,
+                });
+            }
+        }
+    }
+    const searchQueries = grounding?.webSearchQueries ?? [];
+    return { answer, sources, searchQueries, cached: false };
+}
+// ── In-session cache ─────────────────────────────────────────────────────────
+const resultCache = new Map();
+function cacheKey(query) {
+    return query.toLowerCase().trim();
+}
+// ── Extension ────────────────────────────────────────────────────────────────
+export default function (pi) {
+    pi.registerTool({
+        name: "google_search",
+        label: "Google Search",
+        description: "Search the web using Google Search via Gemini. " +
+            "Returns an AI-synthesized answer grounded in Google Search results, plus source URLs. " +
+            "Use this when you need current information from the web: recent events, documentation, " +
+            "product details, technical references, news, etc. " +
+            "Requires GEMINI_API_KEY or Google login. Alternative to Brave-based search tools.",
+        promptSnippet: "Search the web via Google Search to get current information with sources",
+        promptGuidelines: [
+            "Use google_search when you need up-to-date web information that isn't in your training data.",
+            "Be specific with queries for better results, e.g. 'Next.js 15 app router migration guide' not just 'Next.js'.",
+            "The tool returns both an answer and source URLs. Cite sources when sharing results with the user.",
+            "Results are cached per-session, so repeated identical queries are free.",
+            "You can still use fetch_page to read a specific URL if needed after getting results from google_search.",
+        ],
+        parameters: Type.Object({
+            query: Type.String({
+                description: "The search query, e.g. 'latest Node.js LTS version' or 'how to configure Tailwind v4'",
+            }),
+            maxSources: Type.Optional(Type.Number({
+                description: "Maximum number of source URLs to include (default 5, max 10).",
+                minimum: 1,
+                maximum: 10,
+            })),
+        }),
+        async execute(_toolCallId, params, signal, _onUpdate, ctx) {
+            const startTime = Date.now();
+            const maxSources = Math.min(Math.max(params.maxSources ?? 5, 1), 10);
+            // Check for credentials
+            let oauthToken;
+            let projectId;
+            if (!process.env.GEMINI_API_KEY) {
+                const oauthRaw = await ctx.modelRegistry.getApiKeyForProvider("google-gemini-cli");
+                if (oauthRaw) {
+                    try {
+                        const parsed = JSON.parse(oauthRaw);
+                        oauthToken = parsed.token;
+                        projectId = parsed.projectId;
+                    }
+                    catch {
+                        // Fall through to error
+                    }
+                }
+            }
+            if (!process.env.GEMINI_API_KEY && (!oauthToken || !projectId)) {
+                return {
+                    content: [
+                        {
+                            type: "text",
+                            text: "Error: No authentication found for Google Search. Please set GEMINI_API_KEY or log in via Google.\n\nExample: export GEMINI_API_KEY=your_key or use /login google",
+                        },
+                    ],
+                    isError: true,
+                    details: {
+                        query: params.query,
+                        sourceCount: 0,
+                        cached: false,
+                        durationMs: Date.now() - startTime,
+                        error: "auth_error: No credentials set",
+                    },
+                };
+            }
+            // Check cache
+            const key = cacheKey(params.query);
+            if (resultCache.has(key)) {
+                const cached = resultCache.get(key);
+                const output = formatOutput(cached, maxSources);
+                return {
+                    content: [{ type: "text", text: output }],
+                    details: {
+                        query: params.query,
+                        sourceCount: cached.sources.length,
+                        cached: true,
+                        durationMs: Date.now() - startTime,
+                    },
+                };
+            }
+            // Call Gemini with Google Search grounding
+            let result;
+            try {
+                if (process.env.GEMINI_API_KEY) {
+                    const ai = await getClient();
+                    // Add a 30-second timeout to prevent hanging (#1100)
+                    const timeoutController = new AbortController();
+                    const timeoutId = setTimeout(() => timeoutController.abort(), 30_000);
+                    const combinedSignal = signal
+                        ? AbortSignal.any([signal, timeoutController.signal])
+                        : timeoutController.signal;
+                    let response;
+                    try {
+                        response = await ai.models.generateContent({
+                            model: process.env.GEMINI_SEARCH_MODEL || "gemini-2.5-flash",
+                            contents: params.query,
+                            config: {
+                                tools: [{ googleSearch: {} }],
+                                abortSignal: combinedSignal,
+                            },
+                        });
+                    }
+                    finally {
+                        clearTimeout(timeoutId);
+                    }
+                    // Extract answer text
+                    const answer = response.text ?? "";
+                    // Extract grounding metadata
+                    const candidate = response.candidates?.[0];
+                    const grounding = candidate?.groundingMetadata;
+                    // Parse sources from grounding chunks
+                    const sources = [];
+                    const seenTitles = new Set();
+                    if (grounding?.groundingChunks) {
+                        for (const chunk of grounding.groundingChunks) {
+                            if (chunk.web) {
+                                const title = chunk.web.title ?? "Untitled";
+                                // Dedupe by title since URIs are redirect URLs that differ per call
+                                if (seenTitles.has(title))
+                                    continue;
+                                seenTitles.add(title);
+                                // domain field is not available via Gemini API, use title as fallback
+                                // (title is typically the domain name, e.g. "wikipedia.org")
+                                const domain = chunk.web.domain ?? title;
+                                sources.push({
+                                    title,
+                                    uri: chunk.web.uri ?? "",
+                                    domain,
+                                });
+                            }
+                        }
+                    }
+                    // Extract search queries Gemini actually performed
+                    const searchQueries = grounding?.webSearchQueries ?? [];
+                    result = { answer, sources, searchQueries, cached: false };
+                }
+                else {
+                    result = await searchWithOAuth(params.query, oauthToken, projectId, signal);
+                }
+            }
+            catch (err) {
+                const msg = err instanceof Error ? err.message : String(err);
+                let errorType = "api_error";
+                if (msg.includes("401") || msg.includes("UNAUTHENTICATED")) {
+                    errorType = "auth_error";
+                }
+                else if (msg.includes("429") || msg.includes("RESOURCE_EXHAUSTED") || msg.includes("quota")) {
+                    errorType = "rate_limit";
+                }
+                return {
+                    content: [
+                        {
+                            type: "text",
+                            text: `Google Search failed (${errorType}): ${msg}`,
+                        },
+                    ],
+                    isError: true,
+                    details: {
+                        query: params.query,
+                        sourceCount: 0,
+                        cached: false,
+                        durationMs: Date.now() - startTime,
+                        error: `${errorType}: ${msg}`,
+                    },
+                };
+            }
+            // Cache the result
+            resultCache.set(key, result);
+            // Format and truncate output
+            const rawOutput = formatOutput(result, maxSources);
+            const truncation = truncateHead(rawOutput, {
+                maxLines: DEFAULT_MAX_LINES,
+                maxBytes: DEFAULT_MAX_BYTES,
+            });
+            let finalText = truncation.content;
+            if (truncation.truncated) {
+                finalText +=
+                    `\n\n[Truncated: showing ${truncation.outputLines}/${truncation.totalLines} lines` +
+                        ` (${formatSize(truncation.outputBytes)} of ${formatSize(truncation.totalBytes)})]`;
+            }
+            return {
+                content: [{ type: "text", text: finalText }],
+                details: {
+                    query: params.query,
+                    sourceCount: result.sources.length,
+                    cached: false,
+                    durationMs: Date.now() - startTime,
+                },
+            };
+        },
+        renderCall(args, theme) {
+            let text = theme.fg("toolTitle", theme.bold("google_search "));
+            text += theme.fg("accent", `"${args.query}"`);
+            return new Text(text, 0, 0);
+        },
+        renderResult(result, { isPartial, expanded }, theme) {
+            const d = result.details;
+            if (isPartial)
+                return new Text(theme.fg("warning", "Searching Google..."), 0, 0);
+            if (result.isError || d?.error) {
+                return new Text(theme.fg("error", `Error: ${d?.error ?? "unknown"}`), 0, 0);
+            }
+            let text = theme.fg("success", `${d?.sourceCount ?? 0} sources`);
+            text += theme.fg("dim", ` (${d?.durationMs ?? 0}ms)`);
+            if (d?.cached)
+                text += theme.fg("dim", " · cached");
+            if (expanded) {
+                const content = result.content[0];
+                if (content?.type === "text") {
+                    const preview = content.text.split("\n").slice(0, 8).join("\n");
+                    text += "\n\n" + theme.fg("dim", preview);
+                    if (content.text.split("\n").length > 8) {
+                        text += "\n" + theme.fg("muted", "...");
+                    }
+                }
+            }
+            return new Text(text, 0, 0);
+        },
+    });
+    // ── Startup notification ─────────────────────────────────────────────────
+    pi.on("session_start", async (_event, ctx) => {
+        if (process.env.GEMINI_API_KEY)
+            return;
+        const hasOAuth = await ctx.modelRegistry.authStorage.hasAuth("google-gemini-cli");
+        if (!hasOAuth) {
+            ctx.ui.notify("Google Search: No authentication set. Log in via Google or set GEMINI_API_KEY to use google_search.", "warning");
+        }
+    });
+}
+// ── Output formatting ────────────────────────────────────────────────────────
+function formatOutput(result, maxSources) {
+    const lines = [];
+    // Answer
+    if (result.answer) {
+        lines.push(result.answer);
+    }
+    else {
+        lines.push("(No answer text returned from search)");
+    }
+    // Sources
+    if (result.sources.length > 0) {
+        lines.push("");
+        lines.push("Sources:");
+        const sourcesToShow = result.sources.slice(0, maxSources);
+        for (let i = 0; i < sourcesToShow.length; i++) {
+            const s = sourcesToShow[i];
+            lines.push(`[${i + 1}] ${s.title} - ${s.domain}`);
+            lines.push(`    ${s.uri}`);
+        }
+        if (result.sources.length > maxSources) {
+            lines.push(`(${result.sources.length - maxSources} more sources omitted)`);
+        }
+    }
+    else {
+        lines.push("");
+        lines.push("(No source URLs found in grounding metadata)");
+    }
+    // Search queries
+    if (result.searchQueries.length > 0) {
+        lines.push("");
+        lines.push(`Searches performed: ${result.searchQueries.map((q) => `"${q}"`).join(", ")}`);
+    }
+    return lines.join("\n");
+}

package/dist/resources/extensions/gsd/activity-log.js

@@ -0,0 +1,146 @@
+/**
+ * GSD Activity Log — Save raw chat sessions to .gsd/activity/
+ *
+ * Before each context wipe in auto-mode, dumps the full session
+ * as JSONL. No formatting, no truncation, no information loss.
+ * These are debug artifacts — only read when summaries aren't enough.
+ *
+ * Diagnostic extraction is handled by session-forensics.ts.
+ */
+import { writeSync, mkdirSync, readdirSync, unlinkSync, statSync, openSync, closeSync, constants } from "node:fs";
+import { createHash } from "node:crypto";
+import { join } from "node:path";
+import { GSDError, GSD_IO_ERROR } from "./errors.js";
+const SEQ_PREFIX_RE = /^(\d+)-/;
+import { gsdRoot } from "./paths.js";
+const activityLogState = new Map();
+/**
+ * Clear accumulated activity log state (#611).
+ * Call when auto-mode stops to prevent unbounded memory growth
+ * from lastSnapshotKeyByUnit maps accumulating across units.
+ */
+export function clearActivityLogState() {
+    activityLogState.clear();
+}
+function scanNextSequence(activityDir) {
+    let maxSeq = 0;
+    try {
+        for (const f of readdirSync(activityDir)) {
+            const match = f.match(SEQ_PREFIX_RE);
+            if (match)
+                maxSeq = Math.max(maxSeq, parseInt(match[1], 10));
+        }
+    }
+    catch (e) {
+        void e; /* directory not readable — start at 1 */
+        return 1;
+    }
+    return maxSeq + 1;
+}
+function getActivityState(activityDir) {
+    let state = activityLogState.get(activityDir);
+    if (!state) {
+        state = { nextSeq: scanNextSequence(activityDir), lastSnapshotKeyByUnit: new Map() };
+        activityLogState.set(activityDir, state);
+    }
+    return state;
+}
+/**
+ * Build a lightweight dedup key from session entries without serializing
+ * the entire content to a string (#611). Uses entry count + hash of
+ * the last few entries as a fingerprint instead of hashing megabytes.
+ */
+function snapshotKey(unitType, unitId, entries) {
+    const hash = createHash("sha1");
+    hash.update(`${unitType}\0${unitId}\0${entries.length}\0`);
+    // Hash only the last 3 entries as a fingerprint — if the session grew,
+    // the count change alone detects it; if content changed, the tail hash catches it.
+    const tail = entries.slice(-3);
+    for (const entry of tail) {
+        hash.update(JSON.stringify(entry));
+    }
+    return hash.digest("hex");
+}
+function nextActivityFilePath(activityDir, state, unitType, safeUnitId) {
+    // Use O_CREAT | O_EXCL for atomic "create if absent" — no directory scan needed.
+    for (let attempts = 0; attempts < 1000; attempts++) {
+        const seq = String(state.nextSeq).padStart(3, "0");
+        const filePath = join(activityDir, `${seq}-${unitType}-${safeUnitId}.jsonl`);
+        try {
+            const fd = openSync(filePath, constants.O_CREAT | constants.O_EXCL | constants.O_WRONLY);
+            closeSync(fd);
+            return filePath;
+        }
+        catch (err) {
+            if (err?.code === "EEXIST") {
+                state.nextSeq++;
+                continue;
+            }
+            throw err;
+        }
+    }
+    // Fallback: should never reach here in practice
+    throw new GSDError(GSD_IO_ERROR, `Failed to find available activity log sequence in ${activityDir}`);
+}
+export function saveActivityLog(ctx, basePath, unitType, unitId) {
+    try {
+        const entries = ctx.sessionManager.getEntries();
+        if (!entries || entries.length === 0)
+            return null;
+        const activityDir = join(gsdRoot(basePath), "activity");
+        mkdirSync(activityDir, { recursive: true });
+        const safeUnitId = unitId.replace(/\//g, "-");
+        const state = getActivityState(activityDir);
+        const unitKey = `${unitType}\0${safeUnitId}`;
+        // Use lightweight fingerprint instead of serializing all entries (#611)
+        const key = snapshotKey(unitType, safeUnitId, entries);
+        if (state.lastSnapshotKeyByUnit.get(unitKey) === key)
+            return null;
+        const filePath = nextActivityFilePath(activityDir, state, unitType, safeUnitId);
+        // Stream entries to disk line-by-line instead of building one massive string (#611).
+        // For large sessions, the single-string approach allocated hundreds of MB.
+        const fd = openSync(filePath, "w");
+        try {
+            for (const entry of entries) {
+                writeSync(fd, JSON.stringify(entry) + "\n");
+            }
+        }
+        finally {
+            closeSync(fd);
+        }
+        state.nextSeq += 1;
+        state.lastSnapshotKeyByUnit.set(unitKey, key);
+        return filePath;
+    }
+    catch (e) {
+        // Don't let logging failures break auto-mode
+        void e;
+        return null;
+    }
+}
+export function pruneActivityLogs(activityDir, retentionDays) {
+    try {
+        const files = readdirSync(activityDir);
+        const entries = [];
+        for (const f of files) {
+            const match = f.match(SEQ_PREFIX_RE);
+            if (match)
+                entries.push({ seq: parseInt(match[1], 10), filePath: join(activityDir, f) });
+        }
+        if (entries.length === 0)
+            return;
+        const maxSeq = Math.max(...entries.map(e => e.seq));
+        const cutoff = Date.now() - retentionDays * 86_400_000;
+        for (const entry of entries) {
+            if (entry.seq === maxSeq)
+                continue; // always preserve highest-seq
+            try {
+                const mtime = statSync(entry.filePath).mtimeMs;
+                if (Math.floor(mtime) <= cutoff)
+                    unlinkSync(entry.filePath);
+            }
+            catch { /* file vanished or stat failed — skip */ }
+        }
+    }
+    catch { /* empty dir or readdirSync failure — skip */ }
+}

package/dist/resources/extensions/gsd/atomic-write.js

@@ -0,0 +1,38 @@
+import { writeFileSync, renameSync, unlinkSync, mkdirSync, promises as fs } from "node:fs";
+import { dirname } from "node:path";
+import { randomBytes } from "node:crypto";
+/**
+ * Atomically writes content to a file by writing to a temp file first,
+ * then renaming. Prevents partial/corrupt files on crash.
+ */
+export function atomicWriteSync(filePath, content, encoding = "utf-8") {
+    mkdirSync(dirname(filePath), { recursive: true });
+    const tmpPath = filePath + `.tmp.${randomBytes(4).toString("hex")}`;
+    writeFileSync(tmpPath, content, encoding);
+    try {
+        renameSync(tmpPath, filePath);
+    }
+    catch (err) {
+        try {
+            unlinkSync(tmpPath);
+        }
+        catch { /* orphan cleanup best-effort */ }
+        throw err;
+    }
+}
+/**
+ * Async variant of atomicWriteSync. Atomically writes content to a file
+ * by writing to a temp file first, then renaming.
+ */
+export async function atomicWriteAsync(filePath, content, encoding = "utf-8") {
+    await fs.mkdir(dirname(filePath), { recursive: true });
+    const tmpPath = filePath + `.tmp.${randomBytes(4).toString("hex")}`;
+    await fs.writeFile(tmpPath, content, encoding);
+    try {
+        await fs.rename(tmpPath, filePath);
+    }
+    catch (err) {
+        await fs.unlink(tmpPath).catch(() => { });
+        throw err;
+    }
+}