@columenlabs/lumen-ai 0.1.208 → 0.1.209

package/README.md

@@ -103,7 +103,7 @@ app.all('/api/assistant/*', (c) => routes(c.req.raw));
 - **Hand off and advice.** The `escalate` tool hands one task to a higher tier (the person sets next, any or only when asked); Anthropic's advisor tool lets a higher tier send a plan, at most two an answer.
 - **Chats** run on the agent (`./agent`, `LumenAgent`, below): the app routes `POST /chat` and each chat's reads, cards and Undo to its own agent, and `assistantRoutes` answers `POST /chat` with 410 `moved`. Chats saved by the old turn loop stay readable (`GET /chats`, `GET|DELETE /chats/:id`) until `assistant.purge()` drops them 30 days after their last message.
 - **Screen moves**: while the person's Settings ▸ Screen is Move (the default), the panel sends its page map (`AssistantMessage.screen`: routes, nav pages, Settings sections, cards on screen, rows to open) and `LumenAgent` gives the model a `screen` tool (`./agent` `screen.ts`). Each call's moves (go, settings, chart, open, point) are checked against the map, kept in the chat and streamed as a `screen` event; the panel runs them through `AssistantScreenProvider` (mounted by `LumenShell`) with an Undo chip, and their outcomes go back with the next message (`outcomes`). Nothing is saved by a move.
-- **Background work** an app runs for a person (a score, a summary) is `assistant.complete(caller, { task, system, prompt, images?, maxTokens })`: one answer on the task's tier (`tasks`, so the person picks its model in Settings ▸ Models), no chat and no tools (`images`, base64 with their media type, are read with the prompt: a scanned page, a screenshot), billed to their payer and counted in their spend and allowance like a chat turn. It throws `AssistantError` `not_connected`, `app_cap` or `own_cap` when nobody can pay, so the caller waits and tries later.
+- **Background work** an app runs for a person (a score, a summary) is `assistant.complete(caller, { task, system, prompt, images?, maxTokens })`: one answer on the task's tier (`tasks`, so the person picks its model in Settings ▸ Models), no chat and no tools (`images`, base64 with their media type, are read with the prompt: a scanned page, a screenshot), billed to their payer and counted in their spend and allowance like a chat turn. `maxTokens` (default 1,000) goes up to `COMPLETE_MAX_TOKENS` (32,000, thinking included); the answer is streamed, so a long one keeps its connection, and `cut: true` says it stopped at the cap before it ended (the app can ask again, shorter, or continue). It throws `AssistantError` `not_connected`, `app_cap` or `own_cap` when nobody can pay, so the caller waits and tries later.
 
 ### Commands, `@` and `#`
 

package/dist/assistant/engine.d.ts

@@ -56,6 +56,8 @@ export interface AllowanceNotice {
     request: AllowanceRequest;
     unit: AllowanceUnit;
 }
+/** The largest answer `complete` asks for, tokens (thinking included). Streamed, so minutes are fine. */
+export declare const COMPLETE_MAX_TOKENS = 32000;
 /** A change a write tool made at once, as `undo` gets it. */
 export interface AssistantWrite {
     /** The tool's own name (not the `group__tool` the model calls). */
@@ -292,12 +294,14 @@ export declare class Assistant {
         prompt: string;
         /** Photos read with the prompt (a scanned page, a screenshot). */
         images?: CompleteImage[];
+        /** The answer's token cap, thinking included (default 1,000, at most COMPLETE_MAX_TOKENS). */
         maxTokens?: number;
         signal?: AbortSignal;
     }): Promise<{
         text: string;
         usd: number;
         model: string;
+        cut: boolean;
     }>;
     /** Starts a person's ChatGPT plan sign-in: the address to open in a new tab. */
     startSignin(caller: Caller, provider: AssistantProvider): Promise<{

package/dist/assistant/engine.js

@@ -23,6 +23,8 @@ const DOLLARS = { one: 'dollar', many: 'dollars', usd: 1 };
 const round2 = (n) => Math.round(n * 100) / 100;
 /** A request may ask for at most this many times the starter or monthly cap. */
 const REQUEST_MAX_TIMES = 100;
+/** The largest answer `complete` asks for, tokens (thinking included). Streamed, so minutes are fine. */
+export const COMPLETE_MAX_TOKENS = 32_000;
 /** A turn's failure the person can act on. */
 export class AssistantError extends Error {
     code;
@@ -887,7 +889,10 @@ export class Assistant {
                 cache_read_input_tokens: read,
             }, plan);
         }
-        return text || done?.output_text || '';
+        return {
+            text: text || done?.output_text || '',
+            cut: done?.status === 'incomplete' && done.incomplete_details?.reason === 'max_output_tokens',
+        };
     }
     // ---------------------------------------------------------------- one-shot work
     /**
@@ -904,15 +909,20 @@ export class Assistant {
             throw new AssistantError(waitText(code, s.model.provider, !!this.o.allowance), code, 402);
         }
         const clients = this.clients(paying);
-        const max = Math.min(o.maxTokens ?? 1000, 8000);
+        const max = Math.min(o.maxTokens ?? 1000, COMPLETE_MAX_TOKENS);
         let text = '';
+        // The answer stopped at `max` before it ended (its thinking counts toward the cap too).
+        let cut = false;
         try {
             if ((s.model.provider === 'openai' && clients.openai) ||
                 (s.model.provider === 'xai' && clients.xai)) {
-                text = await this.openaiOnce(clients, s, o.system, o.prompt, max, o.signal, o.images);
+                ({ text, cut } = await this.openaiOnce(clients, s, o.system, o.prompt, max, o.signal, o.images));
             }
             else if (clients.anthropic) {
-                const msg = await clients.anthropic.messages.create({
+                // Streamed and read whole: a long answer (minutes at a high cap) keeps its connection alive,
+                // and the SDK refuses a long answer as one unstreamed request.
+                const msg = await clients.anthropic.messages
+                    .stream({
                     model: s.model.id,
                     max_tokens: max,
                     system: o.system,
@@ -935,8 +945,10 @@ export class Assistant {
                         },
                     ],
                     ...effortParams(s.effort),
-                }, o.signal ? { signal: o.signal } : undefined);
+                }, o.signal ? { signal: o.signal } : undefined)
+                    .finalMessage();
                 s.meter.add(msg.usage);
+                cut = msg.stop_reason === 'max_tokens';
                 text = msg.content
                     .filter((b) => b.type === 'text')
                     .map((b) => b.text)
@@ -953,7 +965,7 @@ export class Assistant {
         finally {
             await this.state.addSpend(caller.id, paying.payer, [s.meter.part()]);
         }
-        return { text, usd: s.meter.usd, model: s.model.id };
+        return { text, usd: s.meter.usd, model: s.model.id, cut };
     }
     // ---------------------------------------------------------------- keys
     /** Starts a person's ChatGPT plan sign-in: the address to open in a new tab. */

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@columenlabs/lumen-ai",
-  "version": "0.1.208",
+  "version": "0.1.209",
   "description": "Lumen's assistant and agent parts: a chat assistant's server (Claude tiers, tools over MCP, the web, hand off, spend caps), serving an MCP server from a Worker, read-only tool sets and metered tool calls.",
   "license": "MIT",
   "type": "module",
@@ -52,8 +52,8 @@
     "@ai-sdk/openai": "4.0.108",
     "@anthropic-ai/sdk": "0.133.0",
     "@cloudflare/think": "0.23.0",
-    "@columenlabs/lumen-ops": "0.1.208",
-    "@columenlabs/lumen-ui-types": "0.1.208",
+    "@columenlabs/lumen-ops": "0.1.209",
+    "@columenlabs/lumen-ui-types": "0.1.209",
     "agents": "0.32.0",
     "ai": "7.0.162",
     "just-bash": "3.6.0",