noosphere 0.9.1 → 0.9.3

package/README.md

@@ -694,35 +694,181 @@ const claude = await countTokensAnthropic(messages, ANTHROPIC_KEY, 'claude-sonne
 
 ## Agent Loop & pi-ai Access
 
-Noosphere re-exports the full [pi-ai](https://github.com/nicholasgriffintn/pi-ai) library for direct access to agent loops, tool calling, cost calculation, and streaming APIs.
+Noosphere re-exports the full [pi-ai](https://github.com/mariozechner/pi-ai) library for direct access to agent loops, tool calling, cost calculation, and streaming APIs.
+
+### Preprocessor — Context Compaction
+
+The preprocessor hook runs **before every LLM call** in the agent loop. Use it to manage context window limits — truncate old messages, summarize conversations, or implement sliding window strategies.
 
 ```typescript
-import {
-  agentLoop, calculateCost,
-  piStream, piComplete, piStreamSimple, piCompleteSimple,
-  setApiKey, getApiKey, getPiModel, getPiModels, getPiProviders,
-} from 'noosphere';
+import { agentLoop, getPiModel, setApiKey, countTokensOpenAI } from 'noosphere';
+import type { AgentLoopConfig, AgentContext, PiMessage } from 'noosphere';
 
-// Agent loop with tool calling and preprocessor (compaction hook)
-import type { AgentLoopConfig, AgentContext, AgentTool } from 'noosphere';
+setApiKey('openai', process.env.OPENAI_API_KEY!);
 
 const config: AgentLoopConfig = {
   model: getPiModel('openai', 'gpt-4o'),
-  // Preprocessor runs before each LLM call — use for context compaction
+
+  // Preprocessor runs before each LLM call
   preprocessor: async (messages) => {
-    // Truncate old messages, summarize, etc.
+    // Strategy 1: Simple sliding window — keep last N messages
     if (messages.length > 50) {
-      return messages.slice(-20); // keep last 20
+      return messages.slice(-20);
     }
     return messages;
   },
 };
 
-// Calculate cost before sending
+// Start the agent loop
+const context: AgentContext = {
+  systemPrompt: 'You are a helpful assistant.',
+  messages: [],
+};
+
+const userMessage = {
+  role: 'user' as const,
+  content: 'Hello!',
+  timestamp: Date.now(),
+};
+
+const stream = agentLoop(userMessage, context, config);
+
+for await (const event of stream) {
+  if (event.type === 'message_update' && event.assistantMessageEvent.type === 'text_delta') {
+    process.stdout.write(event.assistantMessageEvent.delta);
+  }
+}
+```
+
+#### Preprocessor Strategies
+
+**Token-aware compaction** — count tokens and trim to fit the context window:
+
+```typescript
+preprocessor: async (messages) => {
+  const model = getPiModel('openai', 'gpt-4o');
+  const maxTokens = model.contextWindow * 0.8; // leave 20% for response
+
+  let totalTokens = 0;
+  const kept: PiMessage[] = [];
+
+  // Keep messages from newest to oldest until we hit the limit
+  for (let i = messages.length - 1; i >= 0; i--) {
+    const msg = messages[i];
+    const content = 'content' in msg && typeof msg.content === 'string' ? msg.content : '';
+    const msgTokens = countTokensOpenAI([{ role: msg.role, content }], 'gpt-4o');
+
+    if (totalTokens + msgTokens > maxTokens) break;
+    totalTokens += msgTokens;
+    kept.unshift(msg);
+  }
+
+  return kept;
+},
+```
+
+**Summarization compaction** — summarize old messages, keep recent ones:
+
+```typescript
+preprocessor: async (messages) => {
+  if (messages.length <= 20) return messages;
+
+  // Summarize the older messages using the LLM itself
+  const oldMessages = messages.slice(0, -10);
+  const recentMessages = messages.slice(-10);
+
+  const summary = await piCompleteSimple(getPiModel('openai', 'gpt-4o-mini'), {
+    systemPrompt: 'Summarize this conversation in 2-3 sentences.',
+    messages: oldMessages,
+  });
+
+  // Replace old messages with a summary message
+  const summaryText = summary.content
+    .filter((c): c is { type: 'text'; text: string } => c.type === 'text')
+    .map((c) => c.text)
+    .join('');
+
+  return [
+    { role: 'user' as const, content: `[Previous conversation summary: ${summaryText}]`, timestamp: Date.now() },
+    ...recentMessages,
+  ];
+},
+```
+
+### Tool Calling
+
+Define tools for the agent loop with typed parameters:
+
+```typescript
+import { Type } from '@sinclair/typebox';
+import type { AgentTool } from 'noosphere';
+
+const weatherTool: AgentTool = {
+  name: 'get_weather',
+  label: 'Get Weather',
+  description: 'Get the current weather for a location',
+  parameters: Type.Object({
+    location: Type.String({ description: 'City name' }),
+  }),
+  execute: async (toolCallId, params) => {
+    const weather = await fetchWeather(params.location);
+    return { output: JSON.stringify(weather), details: weather };
+  },
+};
+
+const context: AgentContext = {
+  systemPrompt: 'You are a helpful assistant with weather access.',
+  messages: [],
+  tools: [weatherTool],
+};
+
+const stream = agentLoop(userMessage, context, config);
+
+for await (const event of stream) {
+  if (event.type === 'tool_execution_start') {
+    console.log(`Calling ${event.toolName}...`);
+  }
+  if (event.type === 'tool_execution_end') {
+    console.log(`Result: ${event.result}`);
+  }
+}
+```
+
+### Cost Calculation
+
+```typescript
+import { calculateCost, getPiModel } from 'noosphere';
+
 const model = getPiModel('openai', 'gpt-4o');
 const usage = { input: 1000, output: 500, cacheRead: 0, cacheWrite: 0 };
 const cost = calculateCost(model, usage);
-console.log(cost.total); // $0.00625
+console.log(cost.total);  // $0.00625
+console.log(cost.input);  // $0.0025
+console.log(cost.output); // $0.00375
+```
+
+### Direct Stream/Complete APIs
+
+```typescript
+import { piComplete, piStream, piCompleteSimple, setApiKey, getPiModel } from 'noosphere';
+
+setApiKey('openai', process.env.OPENAI_API_KEY!);
+const model = getPiModel('openai', 'gpt-4o');
+
+// Simple completion
+const result = await piCompleteSimple(model, {
+  systemPrompt: 'You are helpful.',
+  messages: [{ role: 'user', content: 'Hello!', timestamp: Date.now() }],
+});
+
+// Streaming
+const stream = piStream(model, {
+  messages: [{ role: 'user', content: 'Hello!', timestamp: Date.now() }],
+});
+
+for await (const event of stream) {
+  if (event.type === 'text_delta') process.stdout.write(event.delta);
+}
 ```
 
 ---

package/dist/index.cjs

@@ -2716,6 +2716,36 @@ var OpenAIMediaProvider = class {
   modalities = ["image", "video", "tts", "stt"];
   isLocal = false;
   modelsCache = null;
+  voicesCache = null;
+  /** Auto-fetch available TTS voices by sending an invalid voice and parsing the error. */
+  async fetchVoices() {
+    if (this.voicesCache) return this.voicesCache;
+    try {
+      const res = await fetch(`${OPENAI_API_BASE}/audio/speech`, {
+        method: "POST",
+        headers: {
+          "Content-Type": "application/json",
+          Authorization: `Bearer ${this.apiKey}`
+        },
+        body: JSON.stringify({ model: "tts-1", input: ".", voice: "__discover_voices__" })
+      });
+      if (!res.ok) {
+        const data = await res.json();
+        const msg = data?.error?.message ?? "";
+        const shouldBe = msg.match(/Input should be ([^"]+)/);
+        if (shouldBe) {
+          const voiceList = shouldBe[1].match(/'([a-z]+)'/g);
+          if (voiceList && voiceList.length > 0) {
+            this.voicesCache = voiceList.map((v) => v.replace(/'/g, ""));
+            return this.voicesCache;
+          }
+        }
+      }
+    } catch {
+    }
+    this.voicesCache = [];
+    return this.voicesCache;
+  }
   async ping() {
     try {
       const controller = new AbortController();
@@ -2751,6 +2781,7 @@ var OpenAIMediaProvider = class {
       } finally {
         clearTimeout(timer);
       }
+      const voices = await this.fetchVoices();
       const entries = data?.data ?? [];
       const logo = getProviderLogo("openai");
       const models = [];
@@ -2766,7 +2797,7 @@ var OpenAIMediaProvider = class {
           cost: { price: 0, unit: "per_request" },
           logo,
           description: entry.description,
-          capabilities: this.getCapabilities(entry.id, mod)
+          capabilities: this.getCapabilities(entry.id, mod, voices)
         };
         models.push(info);
       }
@@ -2911,7 +2942,7 @@ var OpenAIMediaProvider = class {
       }
     };
   }
-  getCapabilities(id, modality) {
+  getCapabilities(id, modality, voices) {
     if (modality === "image") {
       return {
         maxWidth: id.startsWith("dall-e-3") ? 1792 : 1024,
@@ -2920,7 +2951,7 @@ var OpenAIMediaProvider = class {
     }
     if (modality === "tts") {
       return {
-        voices: ["alloy", "ash", "coral", "echo", "fable", "onyx", "nova", "sage", "shimmer"]
+        voices: voices && voices.length > 0 ? voices : void 0
       };
     }
     if (modality === "video") {
@@ -2941,18 +2972,34 @@ var OpenAIMediaProvider = class {
 // src/providers/google-media.ts
 var GOOGLE_API_BASE = "https://generativelanguage.googleapis.com/v1beta";
 var FETCH_TIMEOUT_MS6 = 8e3;
-var GOOGLE_TTS_VOICES = [
-  "Aoede",
-  "Charon",
-  "Fenrir",
-  "Kore",
-  "Puck",
-  "Leda",
-  "Orus",
-  "Perseus",
-  "Zephyr",
-  "Callirrhoe"
-];
+async function fetchGoogleVoices(apiKey) {
+  try {
+    const res = await fetch(
+      `${GOOGLE_API_BASE}/models/gemini-2.5-flash-preview-tts:generateContent?key=${apiKey}`,
+      {
+        method: "POST",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({
+          contents: [{ parts: [{ text: "." }] }],
+          generationConfig: {
+            response_modalities: ["AUDIO"],
+            speech_config: { voiceConfig: { prebuiltVoiceConfig: { voiceName: "__discover_voices__" } } }
+          }
+        })
+      }
+    );
+    if (!res.ok) {
+      const data = await res.json();
+      const msg = data?.error?.message ?? "";
+      const match = msg.match(/Allowed voice names are:\s*(.+)/i);
+      if (match) {
+        return match[1].split(",").map((v) => v.trim()).filter(Boolean);
+      }
+    }
+  } catch {
+  }
+  return [];
+}
 function classifyGoogleModel(model) {
   const name = (model.name ?? "").replace("models/", "");
   const methods = model.supportedGenerationMethods ?? [];
@@ -2970,6 +3017,7 @@ var GoogleMediaProvider = class {
   modalities = ["image", "video", "tts"];
   isLocal = false;
   modelsCache = null;
+  voicesCache = null;
   async ping() {
     try {
       const controller = new AbortController();
@@ -3004,6 +3052,9 @@ var GoogleMediaProvider = class {
         clearTimeout(timer);
       }
       const entries = data?.models ?? [];
+      if (!this.voicesCache) {
+        this.voicesCache = await fetchGoogleVoices(this.apiKey);
+      }
       const logo = getProviderLogo("google");
       const models = [];
       for (const entry of entries) {
@@ -3020,7 +3071,7 @@ var GoogleMediaProvider = class {
           cost: { price: 0, unit: modality2 === "video" ? "per_video" : "per_image" },
           logo,
           description: entry.description,
-          capabilities: modality2 === "video" ? { maxDuration: 8, supportsStreaming: false } : modality2 === "tts" ? { voices: GOOGLE_TTS_VOICES } : void 0
+          capabilities: modality2 === "video" ? { maxDuration: 8, supportsStreaming: false } : modality2 === "tts" ? { voices: this.voicesCache && this.voicesCache.length > 0 ? this.voicesCache : void 0 } : void 0
         };
         models.push(info);
       }