bump 1.4.0

package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@ztimson/ai-utils",
-	"version": "1.3.2",
+	"version": "1.4.0",
 	"description": "AI Utility library",
 	"author": "Zak Timson",
 	"license": "MIT",

src/antrhopic.ts

@@ -1,16 +1,27 @@
 import {Anthropic as anthropic} from '@anthropic-ai/sdk';
-import {findByProp, objectMap, JSONSanitize, JSONAttemptParse} from '@ztimson/utils';
+import {findByProp, objectMap, JSONSanitize, JSONAttemptParse, makeArray} from '@ztimson/utils';
 import {AbortablePromise, Ai} from './ai.ts';
 import {LLMMessage, LLMRequest} from './llm.ts';
 import {LLMProvider} from './provider.ts';
+import {TokenPool} from './token-pool.ts';
 import {convertSchema} from './tools.ts';
 
 export class Anthropic extends LLMProvider {
-	client!: anthropic;
+	private clients = new Map<string, anthropic>();
+	tokenPool!: TokenPool;
 
-	constructor(public readonly ai: Ai, public readonly apiToken: string, public model: string) {
+	constructor(public readonly ai: Ai, public readonly apiToken: string | string[], public model: string) {
 		super();
-		this.client = new anthropic({apiKey: apiToken});
+		this.tokenPool = new TokenPool(...makeArray(apiToken).filter(Boolean));
+	}
+
+	private getClient(token: string): anthropic {
+		let client = this.clients.get(token);
+		if(!client) {
+			client = new anthropic({apiKey: token});
+			this.clients.set(token, client);
+		}
+		return client;
 	}
 
 	private toStandard(history: any[]): LLMMessage[] {
@@ -21,10 +32,10 @@ export class Anthropic extends LLMProvider {
 				messages.push(<any>{timestamp, ...h});
 			} else {
 				const textContent = h.content?.filter((c: any) => c.type == 'text').map((c: any) => c.text).join('\n\n');
-				if(textContent) messages.push({role: h.role, content: textContent, timestamp: timestamp});
+				if(textContent) messages.push({role: h.role, content: textContent, timestamp: timestamp, duration: h.duration, tps: h.tps});
 				h.content.forEach((c: any) => {
 					if(c.type == 'tool_use') {
-						messages.push({role: 'tool', id: c.id, name: c.name, args: c.input, timestamp: c.timestamp, content: undefined});
+						messages.push({role: 'tool', id: c.id, name: c.name, args: c.input, timestamp: h.timestamp, content: undefined, duration: h.duration, tps: h.tps});
 					} else if(c.type == 'tool_result') {
 						const m: any = messages.findLast(m => (<any>m).id == c.tool_use_id);
 						if(m) m[c.is_error ? 'error' : 'content'] = c.content;
@@ -86,17 +97,17 @@ export class Anthropic extends LLMProvider {
 				};
 			}
 
-			let resp: any, hasStreamedText = false, terminal = false;
+			let resp: any, terminal = false, duration = 0, tps = 0;
 			do {
 				requestParams.messages = history.map(({timestamp, ...m}) => m);
-				resp = await this.client.messages.create(requestParams).catch(err => {
+				const callStart = Date.now();
+				resp = await this.tokenPool.run(token => this.getClient(token).messages.create(requestParams)).catch(err => {
 					err.message += `\n\nMessages:\n${JSON.stringify(history, null, 2)}`;
 					throw err;
 				});
 
-				// Streaming mode
+				let usage: any;
 				if(options.stream) {
-					if(hasStreamedText) options.stream({text: '\n\n'});
 					resp.content = [];
 					for await (const chunk of resp) {
 						if(controller.signal.aborted) break;
@@ -110,34 +121,38 @@ export class Anthropic extends LLMProvider {
 							if(chunk.delta.type === 'text_delta') {
 								const text = chunk.delta.text;
 								resp.content.at(-1).text += text;
-								if(text) { hasStreamedText = true; options.stream({text}); }
+								options.stream({text});
 							} else if(chunk.delta.type === 'input_json_delta') {
 								resp.content.at(-1).input += chunk.delta.partial_json;
 							}
 						} else if(chunk.type === 'content_block_stop') {
 							const last = resp.content.at(-1);
 							if(last?.input != null) last.input = last.input ? JSONAttemptParse(last.input, {}) : {};
+						} else if(chunk.type === 'message_delta') {
+							if(chunk.usage) usage = chunk.usage;
 						} else if(chunk.type === 'message_stop') {
 							break;
 						}
 					}
+				} else {
+					usage = resp.usage;
 				}
+				duration = Date.now() - callStart;
+				tps = usage?.output_tokens && duration > 0 ? usage.output_tokens / (duration / 1000) : 0;
 
-				// Run tools
 				const toolCalls = resp.content.filter((c: any) => c.type === 'tool_use');
 				if(toolCalls.length && !controller.signal.aborted) {
-					history.push({role: 'assistant', content: resp.content, timestamp: Date.now()});
+					history.push({role: 'assistant', content: resp.content, timestamp: Date.now(), duration, tps});
 					const results = await Promise.all(toolCalls.map(async (toolCall: any) => {
 						const tool = tools.find(findByProp('name', toolCall.name));
 						if(options.stream) options.stream({tool: toolCall.name});
 						if(!tool) return {tool_use_id: toolCall.id, is_error: true, content: 'Tool not found'};
 						try {
-							// Wrap stream so a tool's `done` ends turn gracefully
 							const toolStream = options.stream && ((chunk: any) => {
 								if(chunk.done) { terminal = true; return; }
 								options.stream!(chunk);
 							});
-							const result = await tool.fn(toolCall.input, toolStream, this.ai);
+							const result = await tool.fn(toolCall.input, toolStream, this.ai, toolCall.id);
 							return {type: 'tool_result', tool_use_id: toolCall.id, content: typeof result == 'object' ? JSONSanitize(result) : result};
 						} catch (err: any) {
 							return {type: 'tool_result', tool_use_id: toolCall.id, is_error: true, content: err?.message || err?.toString() || 'Unknown'};
@@ -150,12 +165,19 @@ export class Anthropic extends LLMProvider {
 
 			if(!terminal) {
 				const textContent = resp.content.filter((c: any) => c.type == 'text').map((c: any) => c.text).join('\n\n');
-				history.push({role: 'assistant', content: textContent, timestamp: Date.now()});
+				history.push({role: 'assistant', content: textContent.trim(), timestamp: Date.now(), duration, tps});
 			}
+
 			history = this.toStandard(history);
-			if(options.stream) options.stream({done: true});
 			if(options.history) options.history.splice(0, options.history.length, ...history);
-			const finalContent = history.at(-1)?.content;
+			if(options.stream) options.stream({done: true});
+
+			const turnStart = history.map(h => h.role).lastIndexOf('user');
+			const finalContent = history.slice(turnStart + 1).reduce((str, h) => {
+				if(h.role === 'assistant') return str + (h.content || '');
+				return str;
+			}, '').trim();
+
 			res(options.schema ? JSONAttemptParse(finalContent, finalContent) : finalContent);
 		}), {abort: () => controller.abort()});
 	}

src/index.ts

@@ -5,5 +5,6 @@ export * from './llm';
 export * from './memory';
 export * from './open-ai';
 export * from './provider';
+export * from './token-pool'
 export * from './tools';
 export * from './vision';

src/llm.ts

@@ -9,8 +9,10 @@ import {dirname, join} from 'path';
 import {spawn} from 'node:child_process';
 import {Memory, MemoryCache, MemoryManager, MemoryOptions} from './memory.ts';
 
-export type AnthropicConfig = {proto: 'anthropic', token: string};
+const MAX_AGENT_DEPTH = 5;
-export type OpenAiConfig = {proto: 'openai', host?: string, token: string};
+
+export type AnthropicConfig = {proto: 'anthropic', token: string | string[]};
+export type OpenAiConfig = {proto: 'openai', host?: string, token: string | string[]};
 
 export type Agent = {
 	name: string;
@@ -22,7 +24,6 @@ export type Agent = {
 	skills?: Skill[] | null;
 	tools?: AiTool[] | null;
 	mcp?: McpServer[] | null;
-	/** Explicit whitelist of agents this agent may delegate to. Default: none - must opt-in, self is always excluded */
 	agents?: string[] | null;
 }
 
@@ -48,6 +49,10 @@ export type LLMMessage = {
 	error?: undefined | string;
 	/** Timestamp */
 	timestamp?: number;
+	/** Response duration in ms */
+	duration?: number;
+	/** Tokens per second */
+	tps?: number;
 }
 
 export type LLMRequest = {
@@ -101,8 +106,6 @@ export type Skill = {
 	content: string;
 }
 
-const MAX_AGENT_DEPTH = 5;
-
 class LLM {
 	private memoryManager!: MemoryManager;
 
@@ -119,40 +122,33 @@ class LLM {
 		this.memoryManager = new MemoryManager(this);
 	}
 
-	/**
+	private setupAgent(agents: Agent[] = [], allAgents: Agent[], history: LLMMessage[], aborts: (() => void)[], depth = 0, delegateState: {resp: string | null}): AiTool[] {
-	 * Wrap agents as tools. Nested delegation is opt-in only (empty by default, like
-	 * tools/skills/mcp) and an agent can never call itself even if explicitly whitelisted.
-	 * Delegate results are queued in `pending` and spliced into history by `ask()` after
-	 * the provider's own end-of-turn history sync has already run.
-	 */
-	private setupAgent(agents: Agent[] = [], allAgents: Agent[], pending: Map<string, {resp: string, subHistory: LLMMessage[]}[]>, aborts: (() => void)[], depth = 0): AiTool[] {
 		return agents.map(a => {
 			const toolName = `${a.delegate ? '' : 'sub'}agent_${snakeCase(a.name)}`;
 			return {
 				name: toolName,
 				description: `${a.delegate ? 'Delegate to ' : ''}Subagent: ${a.description || a.name}`,
-				args: {
+				args: <any>(a.delegate ? {} : {
 					context: {type: 'string', description: 'Summary of related messages, samples, files, etc...', required: true},
 					instructions: {type: 'string', description: 'Detailed instructions for subagent to complete', required: true},
-				},
+				}),
-				fn: async (args: any, stream: any) => {
+				fn: async (args: any, stream: any, ai: any, id?: string) => {
 					if(depth >= MAX_AGENT_DEPTH) return 'Max agent delegation depth exceeded';
 
-					const subHistory: LLMMessage[] = [];
 					// Opt-in only, self always excluded regardless of whitelist
 					const nested = (a.agents || [])
 						.map(name => allAgents.find(x => x.name === name))
 						.filter((x): x is Agent => !!x && x.name !== a.name);
 
-					const request = this.ask(`${args.instructions}${args.context ? `\n\n<context>${args.context}</context>` : ''}`, {
+					const request = this.ask(a.delegate ? '' : `${args.instructions}${args.context ? `\n\n<context>${args.context}</context>` : ''}`, {
-						system: `You are a specialized subagent. ${a.delegate ? 'Your output streams directly to the user for the remainder of this turn.' : 'You are wrapped in a tool call that will be analysis by an LLM'}
+						system: `You are a specialized subagent. ${a.delegate ? 'Your output streams directly to the user for the remainder of this turn. You are mid conversation - dispense with greetings.' : 'You are wrapped in a tool call that will be analysis by an LLM - dispense with conversation'}
 As a subagent, focus on executing your task completely using available tools and returning only the final result - no commentary, questions, or dialogue.
 
 ${a.system}`,
 						model: a.model || undefined,
 						temperature: a.temperature,
 						stream: a.delegate ? stream : undefined,
-						history: subHistory,
+						history: a.delegate ? history : [],
 						mcp: a.mcp || undefined,
 						skills: a.skills || undefined,
 						tools: a.tools || undefined,
@@ -163,8 +159,7 @@ ${a.system}`,
 					const resp = await request;
 
 					if(a.delegate) {
-						if(!pending.has(toolName)) pending.set(toolName, []);
+						delegateState.resp = resp;
-						pending.get(toolName)!.push({resp, subHistory});
 						return '';
 					}
 					return resp;
@@ -231,6 +226,20 @@ ${a.system}`,
 		}
 	}
 
+	private wrapToolTiming(tools: AiTool[], timings: Map<string, {duration: number, tps: number}>): AiTool[] {
+		return tools.map(t => ({
+			...t,
+			fn: async (args: any, stream: any, ai: any, id?: string) => {
+				const start = Date.now();
+				const result = await t.fn(args, stream, ai, id);
+				const duration = Date.now() - start;
+				const tps = duration > 0 ? this.estimateTokens(result) / (duration / 1000) : 0;
+				if(id) timings.set(id, {duration, tps});
+				return result;
+			}
+		}));
+	}
+
 	ask(message: string, options: LLMRequest = {}): AbortablePromise<string> {
 		options = <any>{
 			system: '',
@@ -250,7 +259,10 @@ ${a.system}`,
 			nestedAborts.forEach(a => a());
 		};
 
-		const promise = (async () => {
+		let promise: any;
+		const requestStart = Date.now();
+
+		promise = (async () => {
 			let tools: AiTool[] = options.tools || this.ai.options.llm?.tools || [];
 			const prompts: string[] = [];
 			let history = options.history || [];
@@ -273,8 +285,8 @@ ${a.system}`,
 
 			// Agents
 			const agents = options.agents || this.ai.options?.llm?.agents;
-			const pendingDelegates = new Map<string, {resp: string, subHistory: LLMMessage[]}[]>();
+			const delegateState: {resp: string | null} = {resp: null};
-			if(agents?.length) tools.push(...this.setupAgent(agents, agents, pendingDelegates, nestedAborts, options._agentDepth || 0));
+			if(agents?.length) tools.push(...this.setupAgent(agents, agents, history, nestedAborts, options._agentDepth || 0, delegateState));
 
 			// Memory
 			const mem = MemoryManager.normalize(options.memory);
@@ -316,39 +328,36 @@ Also relevant but not preloaded (use \`memory_recall\`): ${listed.map(r => r.nam
 
 			if(aborted) throw Object.assign(new Error('Aborted'), {name: 'AbortError'});
 
+			const toolTimings = new Map<string, {duration: number, tps: number}>();
+			tools = this.wrapToolTiming(tools, toolTimings);
+
+			if(aborted) throw Object.assign(new Error('Aborted'), {name: 'AbortError'});
+
 			prompts.unshift(options.system || this.ai.options.llm?.system || '');
 			request = this.models[m].ask(message, {...options, tools, system: prompts.filter(Boolean).join('\n\n')});
 			let resp = await request;
 
-			// Spice delegated agents response into history
+			// Capture meta (duration / tps)
-			let lastDelegateResp: string | null = null;
+			for(const h of history) {
-			if(pendingDelegates.size) {
+				if(h.role === 'tool' && toolTimings.has(h.id)) Object.assign(h, toolTimings.get(h.id));
-				for(let i = 0; i < history.length; i++) {
-					const h = history[i];
-					if(h.role !== 'tool' || h.content !== '') continue;
-					const queue = pendingDelegates.get(h.name);
-					if(!queue?.length) continue;
-					const {resp: delegateResp, subHistory} = queue.shift()!;
-					const insert: LLMMessage[] = [...subHistory.filter(sh => sh.role === 'tool'), {role: 'assistant', content: delegateResp, timestamp: Date.now()}];
-					history.splice(i + 1, 0, ...insert);
-					lastDelegateResp = delegateResp;
-					i += insert.length;
-				}
 			}
 
-			// If the orchestrator added no commentary of its own, its answer IS the delegate's answer
+			if(typeof resp === 'string' && !resp.trim() && delegateState.resp !== null) resp = delegateState.resp;
-			if(typeof resp === 'string' && !resp.trim() && lastDelegateResp !== null) resp = lastDelegateResp;
 
-			// Trim memory injections from history
 			if(mem?.tool) history.splice(0, history.length, ...history.filter(h => h.role !== 'tool' || h.name !== 'memory_recall'));
-
-			// Auto-memorize before compressing
 			if(options.compress && this.estimateTokens(history) >= options.compress.max) {
 				if(mem?.update) await this.memoryManager.memorize(history, mem.memory, {model: options.memoryModel || this.defaultModel, ...options});
 				const compressed = await this.compressHistory(history, options.compress.max, options.compress.min, options);
 				if(options.history) options.history.splice(0, options.history.length, ...compressed);
 			}
 
+			const requestDuration = Date.now() - requestStart;
+			const totalTokens = history
+				.filter((h: any) => h.role === 'assistant' && h.duration && h.tps)
+				.reduce((sum: number, h: any) => sum + h.tps * (h.duration / 1000), 0);
+			const requestTps = requestDuration > 0 ? totalTokens / (requestDuration / 1000) : 0;
+			Object.assign(promise, {duration: requestDuration, tps: requestTps});
+
 			return resp;
 		})();
 

src/memory.ts

@@ -2,72 +2,15 @@ import {LLMRequest, LLMMessage} from './llm.ts';
 import {AiTool} from './tools.ts';
 import {KDPoint, KDTree} from './kd-tree.ts';
 
-export function buildMemoryGraph(memories: Memory[] | MemoryCache): MemoryNode[] {
+const FACTS_HEADING = '## Facts';
-	const mems = memories instanceof MemoryCache ? memories.memories : memories;
-	const nameSet = new Set(mems.map(m => m.name));
-	const ghosts = new Set<string>();
 
-	const nodes: MemoryNode[] = mems.map(m => {
+const GENERIC_TEMPLATE = `# {{Title}}
-		const {links, backlinks} = extractMetadata(m.content);
-		return {
-			name: m.name,
-			missing: false,
-			links,
-			backlinks,
-		};
-	});
 
-	for (const node of nodes) {
+## Summary
-		for (const link of node.links) {
-			if (!nameSet.has(link)) ghosts.add(link);
-		}
-	}
 
-	return [
+## Details
-		...nodes,
-		...[...ghosts].map(name => ({
-			name,
-			missing: true,
-			links: [],
-			backlinks: nodes
-				.filter(n => n.links.includes(name))
-				.map(n => n.name),
-		}))
-	];
-}
 
-export function renderMemoryGraph(nodes) {
+## Related`;
-	if (!nodes.length) return 'No memories yet.';
-
-	const groups = new Map();
-	for (const node of nodes) {
-		const [prefix, ...rest] = node.name.split('/');
-		const group = rest.length ? prefix : 'Root';
-		const label = rest.length ? rest.join('/') : node.name;
-		if (!groups.has(group)) groups.set(group, []);
-		groups.get(group).push({...node, label});
-	}
-
-	const ghostCount = nodes.filter(n => n.missing).length;
-	const lines = [`Memory Graph (${nodes.length} nodes, ${ghostCount} ghost${ghostCount === 1 ? '' : 's'})`, ''];
-
-	for (const group of [...groups.keys()].sort()) {
-		const items = groups.get(group).sort((a, b) => a.label.localeCompare(b.label));
-		lines.push(`${group}/`);
-		items.forEach((n, i) => {
-			const last = i === items.length - 1;
-			const branch = last ? '└─' : '├─';
-			const pad = last ? '  ' : '│ ';
-			const tag = n.missing ? ' (ghost)' : '';
-			lines.push(`  ${branch} ${n.label}${tag}`);
-			if (n.links.length) lines.push(`  ${pad}   → ${n.links.join(', ')}`);
-			if (n.backlinks.length) lines.push(`  ${pad}   ← ${n.backlinks.join(', ')}`);
-		});
-		lines.push('');
-	}
-
-	return lines.join('\n').trimEnd();
-}
 
 export class MemoryCache {
 	private tree: KDTree<MemoryRef>;
@@ -144,46 +87,35 @@ export type Memory = {
 	description: string;
 	content: string;
 	embedding: number[];
+	links: string[];
+	backlinks: string[];
 }
 
-export type MemoryRef = {
+type MemoryRef = {
 	name: string;
 	description: string;
 }
 
-export type FactBucket = {
+type FactBucket = {
 	subject: string;
 	facts: string[];
 }
 
-export type MemoryNode = {
-	name: string;
-	missing: boolean;
-	links: string[];
-	backlinks: string[];
-}
-
 function extractLinks(content: string): string[] {
 	if (!content) return [];
 	const matches = content.matchAll(/\[\[([^\]]+)\]\]/g);
 	return [...new Set([...matches].map(m => m[1].trim()))];
 }
 
-export function extractMetadata(content: string): {links: string[], backlinks: string[]} {
+export function rebuildGraph(memories: Memory[]): void {
-	const match = content.match(/^---\n([\s\S]*?)\n---/);
+	for (const m of memories) m.links = extractLinks(m.content).filter(l => l !== m.name);
-	if (!match) return {links: [], backlinks: []};
+	for (const m of memories) m.backlinks = [];
-
+	for (const m of memories) {
-	const fm = match[1];
+		for (const link of m.links) {
-	const getList = (key: string): string[] => {
+			const target = memories.find(t => t.name === link);
-		const m = fm.match(new RegExp(`^${key}:\\s*\\[(.*)\\]$`, 'm'));
+			if (target) target.backlinks.push(m.name);
-		if (!m || !m[1].trim()) return [];
+		}
-		return m[1].split(',').map(s => s.trim().replace(/^"|"$/g, '')).filter(Boolean);
+	}
-	};
-
-	return {
-		links: getList('links'),
-		backlinks: getList('backlinks'),
-	};
 }
 
 function dedupeFacts(facts: string[]): string[] {
@@ -214,23 +146,11 @@ function getWeekMonday(date: Date = new Date()): string {
 	return d.toISOString().slice(0, 10);
 }
 
-function getWeekSunday(monday: string): string {
-	const d = new Date(`${monday}T00:00:00Z`);
-	d.setUTCDate(d.getUTCDate() + 6);
-	return d.toISOString().slice(0, 10);
-}
-
 export class MemoryManager {
 	private recentlyTouched = new Map<string, number>();
 
-	private pendingMemorizations = new Map<string, {
-		memories: Memory[] | MemoryCache,
-		tempMemoryName: string,
-		timestamp: number,
-	}>();
-
 	private queues = new Map<string, {
-		pending: string[],
+		dirty: boolean,
 		request: {abort?: () => void} | null,
 		task: Promise<void>,
 	}>();
@@ -243,7 +163,7 @@ export class MemoryManager {
 				name: {type: 'string', description: 'Exact memory name', required: true},
 			},
 			fn: (args: any) => {
-				const mems = memories instanceof MemoryCache ? memories.memories : memories;
+				const mems = this.unwrap(memories);
 				const mem = mems.find(m => m.name === args.name);
 				if (!mem) return 'Document not found';
 				this.touch(mem.name);
@@ -272,110 +192,58 @@ export class MemoryManager {
 		return raw ? {memory: <Memory[] | MemoryCache>m, inject: true, tool: true, update: true} : {inject: true, tool: true, update: true, ...m};
 	}
 
-	private async createTempMemory(conversation: string): Promise<Memory> {
+	private unwrap(memories: Memory[] | MemoryCache): Memory[] {
-		const timestamp = Date.now();
+		return memories instanceof MemoryCache ? memories.memories : memories;
-		const content = `---
-name: _temp_${timestamp}
-description: Temporary memory - processing in background
-tags: [_temporary]
-links: []
-backlinks: []
-modified: ${new Date().toISOString()}
----
-
-# Recent Conversation (Processing)
-
-${conversation}`;
-		const [e] = await this.llm.embedding(content);
-		return {
-			name: `_temp_${timestamp}`,
-			description: 'Temporary memory - processing in background',
-			content,
-			embedding: e?.embedding || [],
-		};
 	}
 
-	private applyHeader(content: string, header: string): string {
+	private sync(memories: Memory[] | MemoryCache): void {
-		return `${header}\n\n${this.stripHeader(content)}`;
+		if (memories instanceof MemoryCache) memories.rebuild();
 	}
 
-	private async backgroundMemorization(conversation: string, memories: Memory[] | MemoryCache, options: LLMRequest, tempName: string): Promise<void> {
+	private parseFrontmatter(content: string): {fm: Map<string, string>, body: string} {
-		const mem = memories instanceof MemoryCache ? memories.memories : memories;
+		const match = content.match(/^---\n([\s\S]*?)\n---\n?([\s\S]*)$/);
-		const monday = getWeekMonday();
+		if (!match) return {fm: new Map(), body: content};
-		const sunday = getWeekSunday(monday);
+		const fm = new Map<string, string>();
-		const buckets = await this.factAgent(conversation, mem, options, monday);
+		for (const line of match[1].split('\n')) {
-		if(!buckets.length) return;
+			const i = line.indexOf(':');
-		const jobs = [...buckets].map(({subject, facts}) => {
+			if (i === -1) continue;
-			let node = mem.find(m => m.name === subject);
+			fm.set(line.slice(0, i).trim(), line.slice(i + 1).trim());
-			if(!node) {
-				node = {name: subject, description: '', content: '', embedding: [],};
-				mem.push(node);
 		}
-			const week = subject.startsWith('Journal/') ? {monday, sunday} : undefined;
+		return {fm, body: match[2]};
-			return this.enqueue(node, facts, mem, options, tempName, week);
-		});
-		await Promise.all(jobs);
 	}
 
-	private buildHeader(node: Memory, week?: {monday: string, sunday: string}, links: string[] = [], backlinks: string[] = []): string {
+	private writeFrontmatter(fm: Map<string, string>, body: string): string {
-		const tags = node.name.split('/')[0]?.toLowerCase();
+		const lines = [...fm.entries()].map(([k, v]) => `${k}: ${v}`);
-		const lines = [
+		return `---\n${lines.join('\n')}\n---\n\n${body.trimStart()}`;
-			'---',
-			`name: ${node.name}`,
-			`description: ${node.description || ''}`,
-			tags ? `tags: [${tags}]` : '',
-			links.length ? `links: [${links.map(l => `"${l}"`).join(', ')}]` : 'links: []',
-			backlinks.length ? `backlinks: [${backlinks.map(l => `"${l}"`).join(', ')}]` : 'backlinks: []',
-			week ? `week: ${week.monday} – ${week.sunday}` : '',
-			`modified: ${new Date().toISOString()}`,
-			'---',
-		].filter(Boolean);
-		return lines.join('\n');
 	}
 
-	private cosineSearch(query: number[], memories: Memory[], limit: number): MemoryRef[] {
+	private stripHeader(content: string): string {
-		const scored = memories
+		return content.replace(/^---[\s\S]*?\n---\n?/, '').trimStart();
-			.filter(m => m.embedding?.length)
-			.map(m => ({
-				ref: {name: m.name, description: m.description},
-				distance: cosineDistance(query, m.embedding),
-			}))
-			.sort((a, b) => a.distance - b.distance)
-			.slice(0, limit);
-		return scored.map(s => s.ref);
 	}
 
-	/**
+	private touchHeader(node: Memory, body: string): string {
-	 * Coalescing queue: if a doc is already compiling, abort the in-flight run, merge its
+		const {fm} = this.parseFrontmatter(node.content);
-	 * facts with the new ones and restart. Never blocks a pending update, never drops facts.
+		fm.set('name', node.name);
-	 */
+		fm.set('description', node.description || '');
-	private enqueue(node: Memory, facts: string[], memories: Memory[] | MemoryCache, options: LLMRequest, tempName: string, week?: {monday: string, sunday: string}): Promise<void> {
+		fm.set('modified', new Date().toISOString());
-		const key = node.name;
+		return this.writeFrontmatter(fm, body);
-		const existing = this.queues.get(key);
-		if (existing) {
-			existing.pending.push(...facts);
-			existing.request?.abort?.();
-			return existing.task;
 	}
 
-		const entry: {pending: string[], request: {abort?: () => void} | null, task: Promise<void>} = {pending: [...facts], request: null, task: Promise.resolve()};
+	private ensureDoc(node: Memory): void {
-		this.queues.set(key, entry);
+		if (node.content) return;
-		const m = memories instanceof MemoryCache ? memories.memories : memories;
+		const title = node.name.split('/').pop() ?? node.name;
-		entry.task = (async () => {
+		node.content = this.touchHeader(node, `# ${title}\n`);
-			while (entry.pending.length) {
-				const batch = dedupeFacts(entry.pending.splice(0, entry.pending.length));
-				const written = await this.docAgent(node, batch, m, options, tempName, week, entry);
-				if (!written) entry.pending.unshift(...batch);
-			}
-		})().finally(() => {
-			this.queues.delete(key);
-			if(!this.queues.size && memories instanceof MemoryCache) memories.rebuild();
-		});
-		return entry.task;
 	}
 
-	private listNodes(memories: Memory[]): MemoryRef[] {
+	private appendFacts(node: Memory, facts: string[]): void {
-		return memories.map(m => ({name: m.name, description: m.description}));
+		this.ensureDoc(node);
+		const body = this.stripHeader(node.content);
+		const bullets = facts.map(f => `- ${f}`).join('\n');
+		const idx = body.indexOf(FACTS_HEADING);
+		const newBody = idx === -1
+			? `${body.trimEnd()}\n\n${FACTS_HEADING}\n${bullets}\n`
+			: `${body.slice(0, idx + FACTS_HEADING.length)}\n${bullets}${body.slice(idx + FACTS_HEADING.length)}`;
+		node.content = this.touchHeader(node, newBody);
 	}
 
 	decay() {
@@ -385,71 +253,27 @@ ${conversation}`;
 		}
 	}
 
-	forget(name: string, memories: Memory[] | MemoryCache): boolean {
+	touch(name: string, ttl = 2) {
-		const mem = memories instanceof MemoryCache ? memories.memories : memories;
+		this.recentlyTouched.set(name, ttl);
-		const idx = mem.findIndex(m => m.name === name);
-		if (idx === -1) return false;
-
-		for (const node of mem) {
-			const {links, backlinks} = extractMetadata(node.content);
-			const newBacklinks = backlinks.filter(b => b !== name);
-			const newLinks = links.filter(l => l !== name);
-
-			if (newBacklinks.length !== backlinks.length || newLinks.length !== links.length) {
-				node.content = this.updateFrontmatter(node.content, {
-					links: newLinks,
-					backlinks: newBacklinks,
-				});
-			}
-		}
-
-		mem.splice(idx, 1);
-
-		if (memories instanceof MemoryCache) memories.rebuild();
-		return true;
 	}
 
 	getTouched(): string[] {
 		return [...this.recentlyTouched.keys()];
 	}
 
-	async memorize(history: LLMMessage[], memories: Memory[] | MemoryCache, options: LLMRequest): Promise<Memory[]> {
+	forget(name: string, memories: Memory[] | MemoryCache): boolean {
-		const conversation = history
+		const mem = this.unwrap(memories);
-			.filter(h => h.role === 'user' || h.role === 'assistant')
+		const idx = mem.findIndex(m => m.name === name);
-			.map(h => `[${h.role}]: ${h.content}`).join('\n\n').trim();
+		if (idx === -1) return false;
-		if(!conversation) return [];
 
-		const trackingId = `${Date.now()}_${Math.random()}`;
+		mem.splice(idx, 1);
-		const tempMemory = await this.createTempMemory(conversation);
+		rebuildGraph(mem);
-		const mem = memories instanceof MemoryCache ? memories.memories : memories;
+		this.sync(memories);
-		mem.push(tempMemory);
+		return true;
-		if (memories instanceof MemoryCache) memories.rebuild();
-		this.pendingMemorizations.set(trackingId, {
-			memories,
-			tempMemoryName: tempMemory.name,
-			timestamp: Date.now(),
-		});
-
-		try {
-			await this.backgroundMemorization(conversation, memories, options, tempMemory.name);
-			const finalMem = memories instanceof MemoryCache ? memories.memories : memories;
-			return finalMem.filter(m => !m.name.startsWith('_temp_'));
-		} finally {
-			const pending = this.pendingMemorizations.get(trackingId);
-			if (pending) {
-				const cleanMem = pending.memories instanceof MemoryCache
-					? pending.memories.memories
-					: pending.memories;
-				const idx = cleanMem.findIndex(m => m.name === pending.tempMemoryName);
-				if (idx !== -1) cleanMem.splice(idx, 1);
-				if (pending.memories instanceof MemoryCache) pending.memories.rebuild();
-			}
-			this.pendingMemorizations.delete(trackingId);
-		}
 	}
 
 	async recollect(query: string, memories: Memory[] | MemoryCache, limit = 5, graphDepth = 1): Promise<Memory[]> {
-		const mem: Memory[] = memories instanceof MemoryCache ? memories.memories : memories;
+		const mem = this.unwrap(memories);
 		if (!mem.length) return [];
 
 		const [e] = await this.llm.embedding(query);
@@ -467,8 +291,7 @@ ${conversation}`;
 				for (const name of frontier) {
 					const node = mem.find(m => m.name === name);
 					if (!node) continue;
-					const {links} = extractMetadata(node.content);
+					for (const link of node.links) {
-					for (const link of links) {
 						if (!found.has(link) && mem.find(m => m.name === link)) {
 							found.add(link);
 							next.push(link);
@@ -486,114 +309,151 @@ ${conversation}`;
 		return ordered.map(n => mem.find(m => m.name === n)!).filter(Boolean);
 	}
 
-	touch(name: string, ttl = 2) {
+	private cosineSearch(query: number[], memories: Memory[], limit: number): MemoryRef[] {
-		this.recentlyTouched.set(name, ttl);
+		const scored = memories
+			.filter(m => m.embedding?.length)
+			.map(m => ({
+				ref: {name: m.name, description: m.description},
+				distance: cosineDistance(query, m.embedding),
+			}))
+			.sort((a, b) => a.distance - b.distance)
+			.slice(0, limit);
+		return scored.map(s => s.ref);
 	}
 
-	private updateFrontmatter(content: string, updates: {links?: string[], backlinks?: string[]}): string {
+	private listNodes(memories: Memory[]): MemoryRef[] {
-		const match = content.match(/^---\n([\s\S]*?)\n---\n\n?([\s\S]*)$/);
+		return memories.map(m => ({name: m.name, description: m.description}));
-		if (!match) return content;
-
-		const [, fm, body] = match;
-		let newFm = fm;
-
-		if (updates.links !== undefined) {
-			const linksList = updates.links.length ? `[${updates.links.map(l => `"${l}"`).join(', ')}]` : '[]';
-			newFm = newFm.replace(/^links:.*$/m, `links: ${linksList}`);
 	}
 
-		if (updates.backlinks !== undefined) {
+	async memorize(history: LLMMessage[], memories: Memory[] | MemoryCache, options: LLMRequest): Promise<Memory[]> {
-			const backlinksList = updates.backlinks.length ? `[${updates.backlinks.map(l => `"${l}"`).join(', ')}]` : '[]';
+		const conversation = history
-			newFm = newFm.replace(/^backlinks:.*$/m, `backlinks: ${backlinksList}`);
+			.filter(h => h.role === 'user' || h.role === 'assistant')
+			.map(h => `[${h.role}]: ${h.content}`).join('\n\n').trim();
+		if (!conversation) return [];
+
+		const uid = `${Date.now()}_${Math.random().toString(36).slice(2)}`;
+		// NOTE: adjust field names below (id/tool_call_id/name) to match your LLMMessage/tool-call schema.
+		const pending = {role: 'tool', name: 'memory_process', id: uid, content: 'Processing…'} as unknown as LLMMessage;
+		history.push(pending);
+
+		const mem = this.unwrap(memories);
+		const buckets = await this.factAgent(conversation, mem, options, getWeekMonday());
+		const touched: Memory[] = [];
+
+		for (const {subject, facts} of buckets) {
+			let node = mem.find(m => m.name === subject);
+			if (!node) {
+				node = {name: subject, description: '', content: '', embedding: [], links: [], backlinks: []};
+				mem.push(node);
+			}
+			this.appendFacts(node, facts);
+			const [e] = await this.llm.embedding(node.content);
+			if (e) node.embedding = e.embedding;
+			this.touch(node.name);
+			touched.push(node);
 		}
 
-		newFm = newFm.replace(/^modified:.*$/m, `modified: ${new Date().toISOString()}`);
+		if (touched.length) {
-
+			rebuildGraph(mem);
-		return `---\n${newFm}\n---\n\n${body}`;
+			this.sync(memories);
+			(pending as any).content = `Saved to ${touched.map(n => `[[${n.name}]]`).join(', ')}`;
+			for (const node of touched) this.reconcile(node, memories, options).catch(() => {});
+		} else {
+			(pending as any).content = 'Nothing worth remembering.';
 		}
 
-	private stripHeader(content: string): string {
+		return touched;
-		return content.replace(/^---[\s\S]*?\n---\n?/, '').trimStart();
 	}
 
-	private async docAgent(node: Memory, facts: string[], memories: Memory[], options: LLMRequest, tempName: string, week: {monday: string, sunday: string} | undefined, entry: {request: {abort?: () => void} | null}): Promise<boolean> {
+	/** Manual/cron entry point. scope 'touched' only reconciles docs with a pending Facts inbox. */
-		const {links: oldLinks} = extractMetadata(node.content);
+	async reconcileVault(memories: Memory[] | MemoryCache, options: LLMRequest, scope: 'touched' | 'all' = 'touched'): Promise<void> {
+		const mem = this.unwrap(memories);
+		const targets = scope === 'all' ? mem : mem.filter(m => m.content.includes(FACTS_HEADING));
+		await Promise.all(targets.map(node => this.reconcile(node, memories, options)));
+		this.sync(memories);
+	}
+
+	/**
+	 * Coalescing queue: if a doc is already reconciling, mark it dirty and abort the in-flight
+	 * request. The loop below always re-reads node.content fresh, so nothing is ever dropped.
+	 */
+	private reconcile(node: Memory, memories: Memory[] | MemoryCache, options: LLMRequest): Promise<void> {
+		const key = node.name;
+		const existing = this.queues.get(key);
+		if (existing) {
+			existing.dirty = true;
+			existing.request?.abort?.();
+			return existing.task;
+		}
+
+		const entry = {dirty: false, request: null, task: Promise.resolve()};
+		this.queues.set(key, entry);
+		const mem = this.unwrap(memories);
+		entry.task = (async () => {
+			do {
+				entry.dirty = false;
+				await this.reconcileDoc(node, mem, options, entry);
+			} while (entry.dirty);
+		})().finally(() => {
+			this.queues.delete(key);
+			rebuildGraph(mem);
+			this.sync(memories);
+		});
+		return entry.task;
+	}
+
+	private async reconcileDoc(node: Memory, memories: Memory[], options: LLMRequest, entry: {request: {abort?: () => void} | null}): Promise<void> {
 		const currentBody = this.stripHeader(node.content);
 		let update;
 		try {
-			for(let i = 0; i < 3 && !update?.content; i++) {
+			for (let i = 0; i < 2 && !update?.content; i++) {
-				const request = this.llm.ask(`New Facts:\n${facts.map(f => `- ${f}`).join('\n')}`, {
+				const request = this.llm.ask(currentBody, {
 					model: options.model,
 					temperature: 0.3,
 					schema: {
 						description: {type: 'string', description: 'One-line description of what this document covers, no formatting or emojis', required: true},
-						content: {type: 'string', description: 'Rewritten document in markdown, without the frontmatter block', required: true},
+						content: {type: 'string', description: 'Rewritten document body in markdown, without the frontmatter block', required: true},
 					},
-					system: `You are a knowledge base editor. Rewrite the current document below so it incorporates the new facts.
+					system: `You are a knowledge base editor maintaining one document in an Obsidian-style vault.
+
+If the document has a "${FACTS_HEADING}" section, integrate every bullet under it into the appropriate part of the document, then remove the "${FACTS_HEADING}" section entirely. If there is no such section, just tidy the document per the rules below.
+
+Structure: follow this generic shape loosely, adapting section names/order to what the content actually needs (e.g. journal-style docs may want a timeline instead of "Details"):
+\`\`\`markdown
+${GENERIC_TEMPLATE}
+\`\`\`
 
 Formatting rules:
-- Use Obsidian-style markdown: # headings, **bold** to add emphasis, __italics__ for titles, terms, etc, bullet & numbered lists for grouped 1D data and tables for 2D data
+- Use Obsidian-style markdown: # headings, **bold** for emphasis, bullet & numbered lists for grouped 1D data, tables for 2D data
 - Link related concepts with [[WikiLink]] notation using full paths like [[People/Sarah]] or [[Projects/Website]]
-- Create links for specific entities (person, place, project, program) and abstract concepts (quantum mechanics, entropy) but skip generics (car, red, dog)
+- Create links for specific entities (person, place, project, program) and abstract concepts, but skip generics (car, red, dog)
 - Keep the document concise, factual, and human-readable
-- Resolve contradictions: the new facts always win — delete the outdated statement entirely, never keep both
+- Resolve contradictions: newer facts always win — delete the outdated statement entirely, never keep both
-- Later facts in the list override earlier ones
 - Do not add frontmatter blocks, filler, preamble, or AI commentary
-${week ? '- This is a weekly journal entry.\n' : ''}
+
-All nodes:
+Other nodes in the vault (link to these instead of duplicating their content):
-${this.listNodes(memories).map(n => n.name).join(', ') || 'none'}
+${this.listNodes(memories).filter(n => n.name !== node.name).map(n => n.name).join(', ') || 'none'}
 
 Current document:
 \`\`\`markdown
 ${currentBody}
-\`\`\``}
+\`\`\``,
-				);
+				});
 				entry.request = request;
 				update = await request;
 			}
 		} catch (err: any) {
-			if (err?.name === 'AbortError') return false;
+			if (err?.name === 'AbortError') return;
 			throw err;
 		} finally {
 			entry.request = null;
 		}
 
-		if(!update?.content) return false;
+		if (!update?.content) return;
-		const newLinks = extractLinks(update.content).filter(l => l !== node.name && l !== tempName);
+		node.description = node.name !== 'People/User' ? update.description : 'All information about the current user';
-		const newLinkSet = new Set(newLinks);
+		node.content = this.touchHeader(node, update.content);
-		const oldLinkSet = new Set(oldLinks);
-
-		for (const added of newLinkSet) {
-			if (!oldLinkSet.has(added)) {
-				const target = memories.find(m => m.name === added);
-				if (target) {
-					const {backlinks} = extractMetadata(target.content);
-					if (!backlinks.includes(node.name)) {
-						target.content = this.updateFrontmatter(target.content, {
-							backlinks: [...backlinks, node.name],
-						});
-					}
-				}
-			}
-		}
-		for (const removed of oldLinkSet) {
-			if (!newLinkSet.has(removed)) {
-				const target = memories.find(m => m.name === removed);
-				if (target) {
-					const {backlinks} = extractMetadata(target.content);
-					target.content = this.updateFrontmatter(target.content, {
-						backlinks: backlinks.filter(b => b !== node.name),
-					});
-				}
-			}
-		}
-
-		const {backlinks} = extractMetadata(node.content);
-		node.description = node.name !== 'Person/User' ? update.description : 'All information about the current user';
-		node.content = this.applyHeader(update.content, this.buildHeader(node, week, newLinks, backlinks));
 		const [e] = await this.llm.embedding(node.content);
 		if (e) node.embedding = e.embedding;
-		return true;
 	}
 
 	private async factAgent(conversation: string, memories: Memory[], options: LLMRequest, weekKey: string): Promise<FactBucket[]> {
@@ -613,13 +473,13 @@ Rules:
 
 When extracting facts, you MUST also decide the exact destination path:
 - Use an existing node name if the facts clearly belong there
-- All information primary about the user should go under "People/User"
+- All information primarily about the user should go under "People/User"
-- When required, create a new path following collection/subject format (e.g., People/Sarah, Projects/Oxide)
+- When required, create a new path following collection/subject format (e.g., People/Sarah, Projects/Oxide) — you are not limited to any fixed list of collections, use whatever fits
 - For journal entries, use "Journal"
 
 Available nodes:
 - Journal
-${this.listNodes(memories).filter(n => !n.name.includes('_temp_') && !n.name.includes('Journal')).map(n => `- ${n.name}: ${n.description}`).join('\n') || 'None yet.'}`,
+${this.listNodes(memories).filter(n => !n.name.includes('Journal')).map(n => `- ${n.name}: ${n.description}`).join('\n') || 'None yet.'}`,
 			tools: [{
 				name: 'facts_extract',
 				description: 'Submit facts with their destination',

src/open-ai.ts

@@ -1,34 +1,47 @@
 import {OpenAI as openAI} from 'openai';
-import {findByProp, objectMap, JSONSanitize, JSONAttemptParse, clean} from '@ztimson/utils';
+import {findByProp, objectMap, JSONSanitize, JSONAttemptParse, clean, makeArray} from '@ztimson/utils';
 import {AbortablePromise, Ai} from './ai.ts';
 import {LLMMessage, LLMRequest} from './llm.ts';
 import {LLMProvider} from './provider.ts';
+import {TokenPool} from './token-pool.ts';
 import {convertSchema} from './tools.ts';
 
 export class OpenAi extends LLMProvider {
-	client!: openAI;
+	tokenPool!: TokenPool;
+	private clients = new Map<string, openAI>();
 
-	constructor(public readonly ai: Ai, public readonly host: string | null, public readonly token: string, public model: string) {
+	constructor(public readonly ai: Ai, public readonly host: string | null, public readonly token: string | string[], public model: string) {
 		super();
-		this.client = new openAI(clean({
+		const tokens = makeArray(token).filter(Boolean);
-			baseURL: host,
+		this.tokenPool = new TokenPool(...(tokens.length ? tokens : [host ? 'ignored' : '']));
-			apiKey: token || (host ? 'ignored' : undefined)
+	}
-		}));
+
+	private getClient(token: string): openAI {
+		let client = this.clients.get(token);
+		if(!client) {
+			client = new openAI(clean({baseURL: this.host, apiKey: token || undefined}));
+			this.clients.set(token, client);
+		}
+		return client;
 	}
 
 	private toStandard(history: any[]): LLMMessage[] {
 		for(let i = 0; i < history.length; i++) {
 			const h = history[i];
 			if(h.role === 'assistant' && h.tool_calls) {
-				const tools = h.tool_calls.map((tc: any) => ({
+				const items: any[] = [];
+				if(h.content) items.push({role: 'assistant', content: h.content, timestamp: h.timestamp, duration: h.duration, tps: h.tps});
+				items.push(...h.tool_calls.map((tc: any) => ({
 					role: 'tool',
 					id: tc.id,
 					name: tc.function.name,
 					args: JSONAttemptParse(tc.function.arguments, {}),
-					timestamp: h.timestamp
+					timestamp: h.timestamp,
-				}));
+					duration: h.duration,
-				history.splice(i, 1, ...tools);
+					tps: h.tps
-				i += tools.length - 1;
+				})));
+				history.splice(i, 1, ...items);
+				i += items.length - 1;
 			} else if(h.role === 'tool') {
 				const record = history.find(h2 => h.tool_call_id == h2.id);
 				if(record) {
@@ -108,25 +121,27 @@ export class OpenAi extends LLMProvider {
 				};
 			}
 
-			let resp: any, hasStreamedText = false, terminal = false;
+			if(options.stream) requestParams.stream_options = {include_usage: true};
+			let resp: any, terminal = false, duration = 0, tps = 0;
 			do {
 				requestParams.messages = history.map(({timestamp, ...m}) => m);
-				resp = await this.client.chat.completions.create(requestParams).catch(err => {
+				const callStart = Date.now();
+				resp = await this.tokenPool.run(token => this.getClient(token).chat.completions.create(requestParams)).catch(err => {
 					err.message += `\n\nMessages:\n${JSON.stringify(history, null, 2)}`;
 					throw err;
 				});
 
+				let usage: any;
 				if(options.stream) {
-					if(hasStreamedText) options.stream({text: '\n\n'});
 					resp.choices = [{message: {role: 'assistant', content: '', tool_calls: [], timestamp: Date.now()}}];
 					for await (const chunk of resp) {
 						if(controller.signal.aborted) break;
-						if(chunk.choices[0].delta.content) {
+						if(chunk.usage) usage = chunk.usage;
-							const text = chunk.choices[0].delta.content;
+						if(chunk.choices[0]?.delta?.content) {
-							resp.choices[0].message.content += text;
+							resp.choices[0].message.content += chunk.choices[0].delta.content;
-							if(text) { hasStreamedText = true; options.stream({text}); }
+							options.stream({text: chunk.choices[0].delta.content});
 						}
-						if(chunk.choices[0].delta.tool_calls) {
+						if(chunk.choices[0]?.delta?.tool_calls) {
 							for(const deltaTC of chunk.choices[0].delta.tool_calls) {
 								const existing = resp.choices[0].message.tool_calls.find(tc => tc.index === deltaTC.index);
 								if(existing) {
@@ -151,24 +166,27 @@ export class OpenAi extends LLMProvider {
 							}
 						}
 					}
+				} else {
+					usage = resp.usage;
 				}
+				duration = Date.now() - callStart;
+				tps = usage?.completion_tokens && duration > 0 ? usage.completion_tokens / (duration / 1000) : 0;
 
 				if(resp.error) throw new Error(resp.error);
 				const toolCalls = resp.choices[0].message.tool_calls || [];
 				if(toolCalls.length && !controller.signal.aborted) {
-					history.push(resp.choices[0].message);
+					history.push({...resp.choices[0].message, duration, tps});
 					const results = await Promise.all(toolCalls.map(async (toolCall: any) => {
 						const tool = tools?.find(findByProp('name', toolCall.function.name));
 						if(options.stream) options.stream({tool: toolCall.function.name});
 						if(!tool) return {role: 'tool', tool_call_id: toolCall.id, content: '{"error": "Tool not found"}', timestamp: Date.now()};
 						try {
 							const args = JSONAttemptParse(toolCall.function.arguments, {});
-							// Wrap stream so a tool's `done` ends turn gracefully
 							const toolStream = options.stream && ((chunk: any) => {
 								if(chunk.done) { terminal = true; return; }
 								options.stream!(chunk);
 							});
-							const result = await tool.fn(args, toolStream, this.ai);
+							const result = await tool.fn(args, toolStream, this.ai, toolCall.id);
 							return {role: 'tool', tool_call_id: toolCall.id, content: typeof result == 'object' ? JSONSanitize(result) : result, timestamp: Date.now()};
 						} catch (err: any) {
 							return {role: 'tool', tool_call_id: toolCall.id, content: JSONSanitize({error: err?.message || err?.toString() || 'Unknown'}), timestamp: Date.now()};
@@ -180,13 +198,20 @@ export class OpenAi extends LLMProvider {
 			} while (!terminal && !controller.signal.aborted && resp.choices?.[0]?.message?.tool_calls?.length);
 
 			if(!terminal) {
-				const textContent = resp.choices[0].message.content?.trim() || '';
+				const textContent = resp.choices[0].message.content || '';
-				history.push({role: 'assistant', content: textContent, timestamp: Date.now()});
+				history.push({role: 'assistant', content: textContent.trim(), timestamp: Date.now(), duration, tps});
 			}
+
 			history = this.toStandard(history);
 			if(options.history) options.history.splice(0, options.history.length, ...history.filter(h => h.role !== 'system'));
 			if(options.stream) options.stream({done: true});
-			const finalContent = history.at(-1)?.content;
+
+			const turnStart = history.map(h => h.role).lastIndexOf('user');
+			const finalContent = history.slice(turnStart + 1).reduce((str, h) => {
+				if(h.role === 'assistant') return str + (h.content || '');
+				return str;
+			}, '').trim();
+
 			res(options.schema ? JSONAttemptParse(finalContent, finalContent) : finalContent);
 		}), {abort: () => controller.abort()});
 	}

src/token-pool.ts

@@ -0,0 +1,65 @@
+const DEFAULT_COOLDOWN = 15 * 60 * 1000;
+
+type TokenState = {
+	token: string;
+	cooldownUntil: number; // 0 = available now
+	lastError?: {code: number, message: string};
+};
+
+export class TokenPoolExhaustedError extends Error {
+	constructor(public tokens: Record<string, {code: number, message: string}>) {
+		super(`All tokens exhausted:\n${Object.entries(tokens).map(([t, e]) => `${t}: [${e.code}] ${e.message}`).join('\n')}`);
+		this.name = 'TokenPoolExhaustedError';
+	}
+}
+
+export class TokenPool {
+	private states: TokenState[];
+
+	constructor(...tokens: string[]) {
+		this.states = tokens.map(token => ({token, cooldownUntil: 0}));
+	}
+
+	private preview(token: string): string {
+		return token.length <= 8 ? '****' : `${token.slice(0, 4)}...${token.slice(-4)}`;
+	}
+
+	/** Anthropic & OpenAI SDKs both attach `status` to thrown errors */
+	private statusCode(err: any): number {
+		return err?.status ?? err?.response?.status ?? err?.statusCode;
+	}
+
+	private retryAfter(err: any): number {
+		const headers = err?.headers || err?.response?.headers;
+		const raw = headers?.get?.('retry-after') ?? headers?.['retry-after'];
+		if(raw) {
+			const seconds = Number(raw);
+			if(!isNaN(seconds)) return Date.now() + seconds * 1000;
+			const date = new Date(raw).getTime();
+			if(!isNaN(date)) return date;
+		}
+		return Date.now() + DEFAULT_COOLDOWN;
+	}
+
+	async run<T>(fn: (token: string) => Promise<T>): Promise<T> {
+		const now = Date.now();
+		for(const state of this.states) {
+			if(state.cooldownUntil > now) continue;
+			try {
+				const result = await fn(state.token);
+				state.cooldownUntil = 0;
+				state.lastError = undefined;
+				return result;
+			} catch(err: any) {
+				const code = this.statusCode(err);
+				if(![401, 403, 429].includes(code)) throw err;
+				state.cooldownUntil = code === 429 ? this.retryAfter(err) : Date.now() + DEFAULT_COOLDOWN;
+				state.lastError = {code, message: err?.message || 'Unknown error'};
+			}
+		}
+
+		const failures: Record<string, {code: number, message: string}> = {};
+		this.states.forEach(s => { if(s.lastError) failures[this.preview(s.token)] = s.lastError; });
+		throw new TokenPoolExhaustedError(failures);
+	}
+}

src/tools.ts

@@ -41,7 +41,7 @@ export type AiTool = {
 	/** Tool arguments */
 	args?: AiToolArg,
 	/** Callback function */
-	fn: (args: any, stream: LLMRequest['stream'], ai: Ai) => any | Promise<any>,
+	fn: (args: any, stream: LLMRequest['stream'], ai: Ai, toolId?: string) => any | Promise<any>,
 };
 
 export function convertSchema(schema: any): any {

tests/llm.spec.ts

@@ -0,0 +1,167 @@
+
+import {describe, it, expect, vi, beforeEach} from 'vitest';
+import LLM from '../src/llm';
+
+const {FakeProvider, providerLog} = vi.hoisted(() => {
+	const providerLog: any[] = [];
+	class FakeProvider {
+		model: string;
+		constructor(...args: any[]) { this.model = args[args.length - 1]; }
+		ask(message: string, opts: any) {
+			let aborted = false;
+			const p = (async () => {
+				const script = (globalThis as any).__scripts?.[this.model];
+				const plan = script ? script(message, opts) : {text: ''};
+				providerLog.push({model: this.model, message, system: opts.system, tools: (opts.tools || []).map((t: any) => t.name)});
+				for (const c of plan.calls || []) {
+					if (aborted) break;
+					const tool = (opts.tools || []).find((t: any) => t.name === c.tool);
+					const id = c.id || `${c.tool}_${Math.random()}`;
+					const content = await tool.fn(c.args, opts.stream, null, id);
+					opts.history.push({role: 'tool', id, name: c.tool, args: c.args, content, timestamp: Date.now()});
+				}
+				const text = plan.text ?? '';
+				if (opts.stream && text) opts.stream({text, done: true});
+				opts.history.push({role: 'assistant', content: text, timestamp: Date.now(), duration: 10, tps: 5});
+				return text;
+			})();
+			return Object.assign(p, {abort: () => { aborted = true; }});
+		}
+	}
+	return {FakeProvider, providerLog};
+});
+
+vi.mock('../src/antrhopic.ts', () => ({Anthropic: FakeProvider}));
+vi.mock('../src/open-ai.ts', () => ({OpenAi: FakeProvider}));
+
+function makeAi(models: any) {
+	return {options: {llm: {models}}} as any;
+}
+
+beforeEach(() => {
+	providerLog.length = 0;
+	(globalThis as any).__scripts = {};
+});
+
+describe('LLM cross-provider interchangeability', () => {
+	it('runs identical tool calls the same way on an anthropic-backed model and an openai-backed model', async () => {
+		const ai = makeAi({
+			claude: {proto: 'anthropic', token: 'x'},
+			gpt: {proto: 'openai', token: 'y', host: 'http://local'},
+		});
+		const llm = new LLM(ai);
+		const calc = {
+			name: 'calc_add',
+			description: 'Add two numbers',
+			args: {a: {type: 'number', required: true}, b: {type: 'number', required: true}},
+			fn: (args: any) => String(args.a + args.b),
+		};
+
+		(globalThis as any).__scripts.claude = () => ({calls: [{tool: 'calc_add', args: {a: 2, b: 3}}], text: 'Result: 5'});
+		(globalThis as any).__scripts.gpt = () => ({calls: [{tool: 'calc_add', args: {a: 2, b: 3}}], text: 'Result: 5'});
+
+		const historyA: any[] = [], historyB: any[] = [];
+		const respA = await llm.ask('add 2 and 3', {model: 'claude', tools: [calc], history: historyA});
+		const respB = await llm.ask('add 2 and 3', {model: 'gpt', tools: [calc], history: historyB});
+
+		expect(respA).toBe('Result: 5');
+		expect(respB).toBe('Result: 5');
+		expect(providerLog.find(l => l.model === 'claude')!.tools).toContain('calc_add');
+		expect(providerLog.find(l => l.model === 'gpt')!.tools).toContain('calc_add');
+
+		// tool timing gets recomputed from real execution regardless of proto
+		for (const h of [historyA.find(h => h.name === 'calc_add'), historyB.find(h => h.name === 'calc_add')]) {
+			expect(h.content).toBe('5');
+			expect(typeof h.duration).toBe('number');
+			expect(typeof h.tps).toBe('number');
+		}
+	});
+
+	it('lets the same shared history flow across model + proto swaps with different system prompts', async () => {
+		const ai = makeAi({
+			claude: {proto: 'anthropic', token: 'x'},
+			gpt: {proto: 'openai', token: 'y', host: 'http://local'},
+		});
+		const llm = new LLM(ai);
+		const history: any[] = [];
+
+		(globalThis as any).__scripts.claude = () => ({text: 'Hi from claude'});
+		(globalThis as any).__scripts.gpt = () => ({text: 'Hi from gpt'});
+
+		const r1 = await llm.ask('hello', {model: 'claude', system: 'You are terse.', history});
+		const r2 = await llm.ask('follow up', {model: 'gpt', system: 'You are verbose.', history});
+
+		expect(r1).toBe('Hi from claude');
+		expect(r2).toBe('Hi from gpt');
+		expect(history.filter(h => h.role === 'assistant').map(h => h.content)).toEqual(['Hi from claude', 'Hi from gpt']);
+		expect(providerLog[0].system).toContain('You are terse.');
+		expect(providerLog[1].system).toContain('You are verbose.');
+	});
+
+	it('exposes MCP tools the same way no matter which proto backs the model', async () => {
+		const ai = makeAi({claude: {proto: 'anthropic', token: 'x'}, gpt: {proto: 'openai', token: 'y', host: 'http://local'}});
+		const llm = new LLM(ai);
+		const mcp = [{name: 'weather', host: 'http://mcp.local'}];
+
+		global.fetch = vi.fn(async (url: string, opts?: any) => {
+			if (url.endsWith('/tools')) {
+				return {json: async () => ({tools: [{name: 'lookup', description: 'Look up weather', inputSchema: {properties: {city: {type: 'string'}}, required: ['city']}}]})} as any;
+			}
+			const body = JSON.parse(opts.body);
+			return {json: async () => ({content: [{text: `Sunny in ${body.arguments.city}`}]})} as any;
+		}) as any;
+
+		for (const model of ['claude', 'gpt']) {
+			(globalThis as any).__scripts[model] = () => ({calls: [{tool: 'weather_lookup', args: {city: 'Rome'}}], text: 'done'});
+			const history: any[] = [];
+			await llm.ask('weather?', {model, mcp, history});
+			expect(history.find(h => h.name === 'weather_lookup')?.content).toBe('Sunny in Rome');
+		}
+	});
+
+	it('exposes and resolves skill documents identically across protos', async () => {
+		const ai = makeAi({claude: {proto: 'anthropic', token: 'x'}, gpt: {proto: 'openai', token: 'y', host: 'http://local'}});
+		const llm = new LLM(ai);
+		const skills = [{name: 'Onboarding', description: 'How to onboard a user', content: 'Step 1...'}];
+
+		for (const model of ['claude', 'gpt']) {
+			(globalThis as any).__scripts[model] = () => ({calls: [{tool: 'skill_read', args: {name: 'Onboarding'}}], text: 'done'});
+			const history: any[] = [];
+			await llm.ask('onboard me', {model, skills, history});
+			expect(history.find(h => h.name === 'skill_read')?.content).toContain('Step 1...');
+		}
+	});
+
+	it('delegate agent mutates the shared history directly and backfills the orchestrator response, across protos', async () => {
+		const ai = makeAi({claude: {proto: 'anthropic', token: 'x'}, gpt: {proto: 'openai', token: 'y', host: 'http://local'}});
+		const llm = new LLM(ai);
+		const history: any[] = [{role: 'user', content: 'research quantum computing'}];
+		const researcher = {name: 'researcher', system: 'You research topics.', delegate: true, model: 'gpt'};
+
+		(globalThis as any).__scripts.claude = () => ({calls: [{tool: 'agent_researcher', args: {}}], text: ''});
+		(globalThis as any).__scripts.gpt = () => ({text: 'Quantum computers use qubits.'});
+
+		const resp = await llm.ask('go', {model: 'claude', agents: [researcher], history});
+
+		expect(resp).toBe('Quantum computers use qubits.');
+		expect(history.some(h => h.role === 'assistant' && h.content === 'Quantum computers use qubits.')).toBe(true);
+		expect(history.find(h => h.name === 'agent_researcher')?.content).toBe('');
+	});
+
+	it('regular (non-delegate) subagent keeps its own isolated history separate from the parent, across protos', async () => {
+		const ai = makeAi({claude: {proto: 'anthropic', token: 'x'}, gpt: {proto: 'openai', token: 'y', host: 'http://local'}});
+		const llm = new LLM(ai);
+		const history: any[] = [];
+		const summarizer = {name: 'summarizer', system: 'You summarize text.', model: 'gpt'};
+
+		(globalThis as any).__scripts.claude = () => ({calls: [{tool: 'subagent_summarizer', args: {context: 'a long article', instructions: 'summarize it'}}], text: 'Summary: short version'});
+		(globalThis as any).__scripts.gpt = () => ({text: 'short version'});
+
+		const resp = await llm.ask('summarize this', {model: 'claude', agents: [summarizer], history});
+
+		expect(resp).toBe('Summary: short version');
+		expect(history.find(h => h.name === 'subagent_summarizer')?.content).toBe('short version');
+		// isolated history - subagent's own assistant turn never leaks into the parent
+		expect(history.some(h => h.role === 'assistant' && h.content === 'short version')).toBe(false);
+	});
+});

tests/memory.spec.ts

@@ -0,0 +1,256 @@
+import {describe, it, expect, vi, beforeEach} from 'vitest';
+import {MemoryManager, MemoryCache, rebuildGraph, Memory} from '../src/memory';
+
+function makeMemory(overrides: Partial<Memory> = {}): Memory {
+	return {
+		name: 'Test/Doc',
+		description: '',
+		content: '',
+		embedding: [],
+		links: [],
+		backlinks: [],
+		...overrides,
+	};
+}
+
+function makeLLM() {
+	return {
+		embedding: vi.fn(async (_text: string) => [{embedding: [1, 0, 0]}]),
+		ask: vi.fn(async () => undefined),
+	};
+}
+
+describe('rebuildGraph', () => {
+	it('extracts [[WikiLinks]] from content, excluding self-links', () => {
+		const a = makeMemory({name: 'A', content: '[[B]] and [[A]] and [[C]]'});
+		const b = makeMemory({name: 'B', content: 'no links here'});
+		const mem = [a, b];
+
+		rebuildGraph(mem);
+
+		expect(a.links).toEqual(['B', 'C']);
+		expect(b.links).toEqual([]);
+	});
+
+	it('computes backlinks only for links that resolve to a real node', () => {
+		const a = makeMemory({name: 'A', content: '[[B]] [[Missing]]'});
+		const b = makeMemory({name: 'B', content: ''});
+		const mem = [a, b];
+
+		rebuildGraph(mem);
+
+		expect(b.backlinks).toEqual(['A']);
+		expect(mem.find(m => m.name === 'Missing')).toBeUndefined();
+	});
+
+	it('resets stale backlinks on every rebuild (no leftover from a removed link)', () => {
+		const a = makeMemory({name: 'A', content: '[[B]]'});
+		const b = makeMemory({name: 'B', content: ''});
+		const mem = [a, b];
+		rebuildGraph(mem);
+		expect(b.backlinks).toEqual(['A']);
+
+		a.content = 'no more links';
+		rebuildGraph(mem);
+		expect(b.backlinks).toEqual([]);
+	});
+});
+
+describe('MemoryCache', () => {
+	it('finds nearest neighbor by embedding via KD-tree search', () => {
+		const close = makeMemory({name: 'Close', embedding: [1, 0, 0]});
+		const far = makeMemory({name: 'Far', embedding: [0, 0, 1]});
+		const cache = new MemoryCache([close, far]);
+
+		const results = cache.search([1, 0, 0], 1);
+
+		expect(results[0].name).toBe('Close');
+	});
+
+	it('rebuilds the tree on add/update/remove', () => {
+		const cache = new MemoryCache([makeMemory({name: 'A', embedding: [1, 0, 0]})]);
+		cache.add(makeMemory({name: 'B', embedding: [0, 1, 0]}));
+		expect(cache.search([0, 1, 0], 1)[0].name).toBe('B');
+
+		cache.remove('B');
+		expect(cache.search([0, 1, 0], 1)[0]?.name).not.toBe('B');
+	});
+});
+
+describe('MemoryManager.forget', () => {
+	it('removes the node and recomputes backlinks for the rest of the graph', () => {
+		const llm = makeLLM();
+		const mgr = new MemoryManager(llm);
+		const a = makeMemory({name: 'A', content: '[[B]]'});
+		const b = makeMemory({name: 'B', content: '[[C]]'});
+		const c = makeMemory({name: 'C', content: ''});
+		const mem = [a, b, c];
+		rebuildGraph(mem);
+		expect(c.backlinks).toEqual(['B']);
+
+		const ok = mgr.forget('B', mem);
+
+		expect(ok).toBe(true);
+		expect(mem.find(m => m.name === 'B')).toBeUndefined();
+		expect(a.links).toEqual(['B']);
+		expect(c.backlinks).toEqual([]);
+	});
+
+	it('returns false for an unknown name', () => {
+		const mgr = new MemoryManager(makeLLM());
+		expect(mgr.forget('Nope', [makeMemory({name: 'A'})])).toBe(false);
+	});
+});
+
+describe('MemoryManager.recollect', () => {
+	it('orders vector matches first, then expands one hop via links', async () => {
+		const llm = makeLLM();
+		llm.embedding.mockResolvedValue([{embedding: [1, 0, 0]}]);
+		const mgr = new MemoryManager(llm);
+
+		const near = makeMemory({name: 'Near', embedding: [1, 0, 0], content: '[[Linked]]'});
+		const linked = makeMemory({name: 'Linked', embedding: [0, 0, 1], content: ''});
+		const far = makeMemory({name: 'Far', embedding: [0, 1, 0], content: ''});
+		const mem = [near, linked, far];
+		rebuildGraph(mem);
+
+		const result = await mgr.recollect('query', mem, 1, 1);
+
+		expect(result.map(r => r.name)).toEqual(['Near', 'Linked']);
+	});
+
+	it('returns [] when there are no memories', async () => {
+		const mgr = new MemoryManager(makeLLM());
+		expect(await mgr.recollect('q', [])).toEqual([]);
+	});
+});
+
+describe('MemoryManager.memorize (fast path)', () => {
+	let llm: ReturnType<typeof makeLLM>;
+	let mgr: MemoryManager;
+
+	beforeEach(() => {
+		llm = makeLLM();
+		mgr = new MemoryManager(llm);
+	});
+
+	it('pushes a pending tool message, then resolves it to links once facts land', async () => {
+		llm.ask.mockImplementation(async (_prompt: string, opts: any) => {
+			if (opts.tools) {
+				opts.tools[0].fn({destination: 'Projects/Oxide', facts: 'Uses a hybrid memory system'});
+				return undefined;
+			}
+			return {description: 'd', content: '# doc'};
+		});
+
+		const history: any[] = [{role: 'user', content: 'we use a hybrid memory system'}];
+		const touched = await mgr.memorize(history, [], {model: 'test'} as any);
+
+		const pending = history.find(h => h.name === 'memory_process');
+		expect(pending).toBeDefined();
+		expect(pending.content).toContain('[[Projects/Oxide]]');
+		expect(touched.map(t => t.name)).toEqual(['Projects/Oxide']);
+	});
+
+	it('creates a new node and appends facts under "## Facts" without calling the doc LLM', async () => {
+		llm.ask.mockImplementation(async (_prompt: string, opts: any) => {
+			if (opts.tools) opts.tools[0].fn({destination: 'People/Sarah', facts: 'Works at Acme, Likes hiking'});
+			return undefined;
+		});
+
+		const mem: Memory[] = [];
+		await mgr.memorize([{role: 'user', content: 'Sarah works at Acme and likes hiking'}] as any, mem, {model: 'test'} as any);
+
+		const node = mem.find(m => m.name === 'People/Sarah')!;
+		expect(node).toBeDefined();
+		expect(node.content).toContain('## Facts');
+		expect(node.content).toContain('- Works at Acme');
+		expect(node.content).toContain('- Likes hiking');
+		// doc reconciler LLM (schema call) should NOT have been awaited synchronously in this fast path assertion
+	});
+
+	it('routes "journal" destination to Journal/{weekMonday}', async () => {
+		llm.ask.mockImplementation(async (_prompt: string, opts: any) => {
+			if (opts.tools) opts.tools[0].fn({destination: 'journal', facts: 'Shipped v1'});
+			return undefined;
+		});
+
+		const mem: Memory[] = [];
+		const touched = await mgr.memorize([{role: 'user', content: 'shipped v1 today'}] as any, mem, {model: 'test'} as any);
+
+		expect(touched[0].name).toMatch(/^Journal\/\d{4}-\d{2}-\d{2}$/);
+	});
+
+	it('reports nothing to remember when no facts are extracted', async () => {
+		llm.ask.mockResolvedValue(undefined); // tools present but fn never called
+
+		const history: any[] = [{role: 'user', content: 'hey'}];
+		const touched = await mgr.memorize(history, [], {model: 'test'} as any);
+
+		expect(touched).toEqual([]);
+		expect(history.find(h => h.name === 'memory_process').content).toBe('Nothing worth remembering.');
+	});
+
+	it('returns [] and does nothing for an empty conversation', async () => {
+		const touched = await mgr.memorize([], [], {model: 'test'} as any);
+		expect(touched).toEqual([]);
+		expect(llm.ask).not.toHaveBeenCalled();
+	});
+});
+
+describe('MemoryManager reconcileVault', () => {
+	it('integrates the "## Facts" section via the doc LLM and removes it', async () => {
+		const llm = makeLLM();
+		llm.ask.mockResolvedValue({description: 'Tidy summary', content: '# Doc\n\nIntegrated fact.'});
+		const mgr = new MemoryManager(llm);
+
+		const node = makeMemory({
+			name: 'Projects/Oxide',
+			content: '---\nname: Projects/Oxide\n---\n\n# Doc\n\n## Facts\n- some raw fact\n',
+		});
+		const mem = [node];
+
+		await mgr.reconcileVault(mem, {model: 'test'} as any, 'all');
+
+		expect(node.content).not.toContain('## Facts');
+		expect(node.content).toContain('Integrated fact.');
+		expect(node.description).toBe('Tidy summary');
+	});
+
+	it('only targets docs with a pending Facts inbox when scope is "touched"', async () => {
+		const llm = makeLLM();
+		llm.ask.mockResolvedValue({description: 'd', content: '# clean'});
+		const mgr = new MemoryManager(llm);
+
+		const dirty = makeMemory({name: 'A', content: '## Facts\n- x'});
+		const clean = makeMemory({name: 'B', content: '# already tidy'});
+		await mgr.reconcileVault([dirty, clean], {model: 'test'} as any, 'touched');
+
+		expect(dirty.content).toContain('# clean'); // rewritten (frontmatter now wraps it)
+		expect(clean.content).toBe('# already tidy'); // untouched, never queued
+	});
+});
+
+describe('MemoryManager reconcile coalescing', () => {
+	it('coalesces a second call while one is in-flight: marks dirty, aborts, reuses the same task promise', () => {
+		const llm = makeLLM();
+		const abort = vi.fn();
+		let calls = 0;
+		llm.ask.mockImplementation(() => {
+			calls++;
+			const pending: any = new Promise(() => {}); // never resolves in this test
+			pending.abort = abort;
+			return pending;
+		});
+		const mgr: any = new MemoryManager(llm);
+		const node = makeMemory({name: 'Q', content: '# Q\n\n## Facts\n- f'});
+		const mem = [node];
+
+		const p1 = mgr.reconcile(node, mem, {model: 'test'});
+		const p2 = mgr.reconcile(node, mem, {model: 'test'});
+
+		expect(p2).toBe(p1); // same in-flight task, not a new queue entry
+		expect(abort).toHaveBeenCalledTimes(1); // second call aborted the in-flight request
+		expect(calls).toBe(1); // no second ask() fired synchronously — it'll rerun via the dirty loop
+	});
+});