Added wikipedia tools

package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@ztimson/ai-utils",
-	"version": "0.8.12",
+	"version": "0.8.16",
 	"description": "AI Utility library",
 	"author": "Zak Timson",
 	"license": "MIT",

src/llm.ts

@@ -358,7 +358,7 @@ class LLM {
 	 * @returns {Promise<{} | {} | RegExpExecArray | null>}
 	 */
 	async json(text: string, schema: string, options?: LLMRequest): Promise<any> {
-		let system =  `Your job is to convert input to JSON. Call \`submit\` exactly once with JSON matching this schema:\n\`\`\`json\n${schema}\n\`\`\``;
+		let system =  `Your job is to convert input to JSON using tool calls. Call the \`submit\` tool at least once with JSON matching this schema:\n\`\`\`json\n${schema}\n\`\`\`\n\nResponses are ignored`;
 		if(options?.system) system += '\n\n' + options.system;
 		return new Promise(async (resolve, reject) => {
 			let done = false;
@@ -376,23 +376,23 @@ class LLM {
 							resolve(json);
 							done = true;
 						} catch { return 'Invalid JSON'; }
-						return 'Done';
+						return 'Saved';
 					}
 				}, ...(options?.tools || [])],
 			});
-			if(!done) reject(`AI failed to create summary: ${resp}`);
+			if(!done) reject(`AI failed to create JSON:\n${resp}`);
 		});
 	}
 
 	/**
 	 * Create a summary of some text
 	 * @param {string} text Text to summarize
-	 * @param {number} tokens Max number of tokens
+	 * @param {number} length Max number of words
 	 * @param options LLM request options
 	 * @returns {Promise<string>} Summary
 	 */
-	async summarize(text: string, tokens: number = 500, options?: LLMRequest): Promise<string | null> {
-		let system =  `Your job is to summarize the users message. Call \`submit\` exactly once with the shortest summary possible that's <= ${tokens} tokens. Output nothing else`;
+	async summarize(text: string, length: number = 500, options?: LLMRequest): Promise<string | null> {
+		let system =  `Your job is to summarize the users message using tool calls. Call the \`submit\` tool at least once with the shortest summary possible that's <= ${length} words. The tool call will respond with the token count. Responses are ignored`;
 		if(options?.system) system += '\n\n' + options.system;
 		return new Promise(async (resolve, reject) => {
 			let done = false;
@@ -403,15 +403,18 @@ class LLM {
 				tools: [{
 					name: 'submit',
 					description: 'Submit summary',
-					args: {summary: {type: 'string', description: 'Summarization', required: true}},
+					args: {summary: {type: 'string', description: 'Text summarization', required: true}},
 					fn: (args) => {
+						if(!args.summary) return 'No summary provided';
+						const count = args.summary.split(' ').length;
+						if(count > length) return `Too long: ${length} words`;
 						done = true;
 						resolve(args.summary || null);
-						return 'Done';
+						return `Saved: ${length} words`;
 					}
 				}, ...(options?.tools || [])],
 			});
-			if(!done) reject(`AI failed to create summary: ${resp}`);
+			if(!done) reject(`AI failed to create summary:\n${resp}`);
 		});
 	}
 }

src/tools.ts

@@ -272,3 +272,99 @@ export const WebSearchTool: AiTool = {
 		return results;
 	}
 }
+
+class WikipediaClient {
+	private ua = 'AiTools-Wikipedia/1.0';
+
+	private async get(url: string): Promise<any> {
+		const resp = await fetch(url, {headers: {'User-Agent': this.ua}});
+		return resp.json();
+	}
+
+	private api(params: Record<string, any>): Promise<any> {
+		const qs = new URLSearchParams({...params, format: 'json', utf8: '1'}).toString();
+		return this.get(`https://en.wikipedia.org/w/api.php?${qs}`);
+	}
+
+	private clean(text: string): string {
+		return text.replace(/\n{3,}/g, '\n\n').replace(/ {2,}/g, ' ').replace(/\[\d+\]/g, '').trim();
+	}
+
+	private truncate(text: string, max: number): string {
+		if(text.length <= max) return text;
+		const cut = text.slice(0, max);
+		const lastPara = cut.lastIndexOf('\n\n');
+		return lastPara > max * 0.7 ? cut.slice(0, lastPara) : cut;
+	}
+
+	private async searchTitles(query: string, limit = 6): Promise<any[]> {
+		const data = await this.api({action: 'query', list: 'search', srsearch: query, srlimit: limit, srprop: 'snippet'});
+		return data.query?.search || [];
+	}
+
+	private async fetchExtract(title: string, intro = false): Promise<string> {
+		const params: any = {action: 'query', prop: 'extracts', titles: title, explaintext: 1, redirects: 1};
+		if(intro) params.exintro = 1;
+		const data = await this.api(params);
+		const page = Object.values(data.query?.pages || {})[0] as any;
+		return this.clean(page?.extract || '');
+	}
+
+	private pageUrl(title: string): string {
+		return `https://en.wikipedia.org/wiki/${encodeURIComponent(title.replace(/ /g, '_'))}`;
+	}
+
+	private stripHtml(text: string): string {
+		return text.replace(/<[^>]+>/g, '');
+	}
+
+	async lookup(query: string, detail: 'intro' | 'full' = 'intro'): Promise<string> {
+		const results = await this.searchTitles(query, 6);
+		if(!results.length) return `❌ No Wikipedia articles found for "${query}"`;
+
+		const title = results[0].title;
+		const url = this.pageUrl(title);
+		const content = await this.fetchExtract(title, detail === 'intro');
+
+		const text = this.truncate(content, detail === 'intro' ? 2000 : 8000);
+		return `## ${title}\n🔗 ${url}\n\n${text}`;
+	}
+
+	async search(query: string): Promise<string> {
+		const results = await this.searchTitles(query, 8);
+		if(!results.length) return `❌ No results for "${query}"`;
+
+		const lines = [`### Search results for "${query}"\n`];
+		for(let i = 0; i < results.length; i++) {
+			const r = results[i];
+			const snippet = this.truncate(this.stripHtml(r.snippet || ''), 150);
+			lines.push(`**${i + 1}. ${r.title}**\n${snippet}\n${this.pageUrl(r.title)}`);
+		}
+		return lines.join('\n\n');
+	}
+}
+
+export const WikipediaLookupTool: AiTool = {
+	name: 'wikipedia_lookup',
+	description: 'Get Wikipedia article content',
+	args: {
+		query: {type: 'string', description: 'Topic or article title', required: true},
+		detail: {type: 'string', description: 'Content level: "intro" (summary, default) or "full" (complete article)', enum: ['intro', 'full'], default: 'intro'}
+	},
+	fn: async (args: {query: string; detail?: 'intro' | 'full'}) => {
+		const wiki = new WikipediaClient();
+		return wiki.lookup(args.query, args.detail || 'intro');
+	}
+};
+
+export const WikipediaSearchTool: AiTool = {
+	name: 'wikipedia_search',
+	description: 'Search Wikipedia for matching articles',
+	args: {
+		query: {type: 'string', description: 'Search terms', required: true}
+	},
+	fn: async (args: {query: string}) => {
+		const wiki = new WikipediaClient();
+		return wiki.search(args.query);
+	}
+};