> ## Documentation Index > Fetch the complete documentation index at: https://docs.thinnest.ai/llms.txt > Use this file to discover all available pages before exploring further. # Built-in Tools > Reference guide for all built-in tools available on thinnestAI — search, scraping, data, productivity, communication, and more. # Built-in Tools thinnestAI comes with dozens of pre-built tools that your agents can use out of the box. Most only require an API key to get started. ## Search Tools Tools for searching the web, academic databases, and knowledge repositories. | Tool | Description | API Key Required | | ------------------------ | ------------------------------------------------------------------------------------------------------------------------------- | :--------------: | | **DuckDuckGo** | Privacy-focused web search. No API key needed — great for getting started. | No | | **Tavily** | AI-optimized search engine built for agents. Returns clean, structured results. | Yes | | **Exa** | Neural search engine that understands meaning, not just keywords. Excellent for research tasks. | Yes | | **SerpAPI** | Google Search results via API. Supports Google, Bing, Yahoo, and more. | Yes | | **Brave Search** | Privacy-focused search engine with its own independent index. High-quality results without tracking. | Yes | | **Web Search (Generic)** | A unified web search interface that works across multiple providers. Good default choice when you don't need a specific engine. | Varies | | **Baidu Search** | China's largest search engine. Essential for agents serving Chinese-language users. | Yes | | **AgentQL** | AI-native search and data extraction. Converts natural language queries into structured web data. | Yes | | **Linkup** | Real-time web search optimized for AI agents. Returns structured, up-to-date results with source attribution. | Yes | | **Wikipedia** | Search and retrieve Wikipedia articles. Useful for factual lookups and definitions. | No | | **PubMed** | Search biomedical and life science literature from the National Library of Medicine. | No | | **ArXiv** | Search academic papers across physics, math, CS, and more. | No | ### Setting Up a Search Tool 1. Go to your agent's **Tools** section. 2. Click **Add Tool** and select a search tool (e.g., Tavily). 3. Enter your API key if required. 4. Save — your agent can now search the web. ```bash theme={null} # Example: Get a Tavily API key TAVILY_API_KEY=tvly-xxxxxxxxxxxxx ``` ## Web Scraping Tools Tools for extracting content from web pages and turning them into clean text. | Tool | Description | API Key Required | | --------------- | ------------------------------------------------------------------------------------------------------------------------------- | :--------------: | | **Firecrawl** | Powerful web scraper that handles JavaScript-rendered pages. Returns clean markdown. | Yes | | **Crawl4AI** | Open-source web crawler optimized for AI. Handles complex sites. | No | | **JinaReader** | Converts any URL into clean, LLM-friendly text. Simple and fast. | Yes | | **Spider** | High-performance web crawler with anti-bot handling. | Yes | | **Newspaper** | Extracts and parses news articles. Pulls title, authors, text, and images. | No | | **Browserbase** | Cloud-hosted headless browsers for scraping JavaScript-heavy sites. Handles authentication, CAPTCHAs, and complex interactions. | Yes | | **Bright Data** | Enterprise web data platform with proxy networks and pre-built collectors for structured data extraction at scale. | Yes | | **Oxylabs** | Premium proxy and scraping infrastructure. Excellent for large-scale, geo-targeted data collection. | Yes | | **ScrapeGraph** | AI-powered scraping that understands page structure. Extracts data using natural language instructions. | Yes | ### When to Use Scraping Tools * **Firecrawl** — Best for JavaScript-heavy sites (SPAs, dashboards). * **JinaReader** — Best for quick, clean text extraction from articles and docs. * **Crawl4AI** — Best when you need a free, self-hosted option. * **Newspaper** — Best specifically for news articles and blog posts. ## Data & Analytics Tools Tools for querying databases, analyzing CSVs, and working with structured data. | Tool | Description | API Key Required | | ----------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :---------------: | | **[Dash (Self-Learning DB)](/docs/tools/dash)** | Self-learning data agent with 6-layer context. Learns from every query, auto-corrects errors, and gets smarter over time. Inspired by OpenAI's in-house data agent. | Connection string | | **DuckDB** | In-process SQL database. Query CSV, Parquet, and JSON files with SQL. | No | | **CSV Tools** | Read, filter, and analyze CSV files directly. | No | | **SQL** | Connect to any SQL database and run queries. | Connection string | | **PostgreSQL** | Direct PostgreSQL database integration for queries and operations. | Connection string | | **BigQuery** | Query Google BigQuery datasets. Great for large-scale analytics. | Yes | | **Neo4j** | Graph database queries and operations. Ideal for relationship-heavy data like org charts, social networks, and knowledge graphs. | Connection string | | **Pandas** | Python-powered data analysis. Transform, filter, aggregate data. | No | ### Dash vs PostgreSQL Tool | Feature | Dash | PostgreSQL | | -------------------- | ------------------------------------------------------- | ----------- | | Schema discovery | Auto-discovers tables, columns, indexes, constraints | Manual | | Business annotations | Per-table/column notes, metric definitions | None | | Self-learning | Learns from successful queries and error corrections | None | | Auto-retry | Diagnoses errors, suggests fixes, retries automatically | None | | Security | Read-only by default, blocks writes | Full access | | Visualization | Suggests chart types for results | None | **Use Dash** for natural language data analysis where the agent queries a database on behalf of users. **Use PostgreSQL** for direct database operations where you need full read/write access. ### Example: Data Analysis Agent Assign the DuckDB and CSV tools to create an agent that can analyze uploaded data: ```bash theme={null} # Agent can answer questions like: # "What were our top 10 customers by revenue last quarter?" # "Show me the trend of signups over the past 6 months" # "Find all orders where the total exceeds $500" ``` ## Finance Tools Tools for accessing financial data, stock prices, and market information. | Tool | Description | API Key Required | | ---------------------- | ------------------------------------------------------------------------------------------------------------------------------- | :--------------: | | **YFinance** | Yahoo Finance data — stock prices, company info, historical data, financials. | No | | **OpenBB** | Professional-grade financial data platform. Covers stocks, crypto, forex, and economics. | Yes | | **Financial Datasets** | Access curated financial datasets including SEC filings, earnings transcripts, institutional holdings, and economic indicators. | Yes | | **Valyu** | Real-time financial data API with market data, company fundamentals, and alternative data signals. | Yes | ### Example: Financial Research Agent ```bash theme={null} # Agent can answer: # "What's Apple's current stock price?" # "Compare Tesla and Ford's P/E ratios" # "Show me Bitcoin's price over the last 30 days" ``` ## AI Generation Tools Tools for generating images, audio, and other media using AI models. | Tool | Description | API Key Required | | ------------------ | ---------------------------------------------------------------------- | :--------------: | | **DALL-E 3** | OpenAI's image generation model. Create images from text descriptions. | Yes (OpenAI) | | **ElevenLabs TTS** | High-quality text-to-speech. Generate natural-sounding audio. | Yes | | **Replicate** | Run open-source AI models (Stable Diffusion, LLaMA, etc.) via API. | Yes | | **FAL** | Fast AI model inference. Image generation, video, and more. | Yes | ## Productivity Tools Tools for project management, task tracking, documentation, and scheduling. | Tool | Description | API Key Required | | -------------- | ------------------------------------------------------------------- | :--------------: | | **GitHub** | Manage repositories, issues, pull requests, and code. | Yes (Token) | | **Linear** | Modern project management. Create and update issues, track sprints. | Yes | | **Todoist** | Task management. Create, complete, and organize tasks and projects. | Yes | | **Confluence** | Atlassian wiki. Search and create documentation pages. | Yes | | **ClickUp** | All-in-one project management. Tasks, docs, goals, and more. | Yes | | **Cal.com** | Open-source scheduling. Book meetings and manage availability. | Yes | ## Communication Tools Tools for messaging, notifications, and meeting management. | Tool | Description | API Key Required | | ------------------- | ----------------------------------------------------------------------- | :--------------: | | **Slack** | Send messages, read channels, manage conversations in Slack workspaces. | Yes (Bot Token) | | **Discord** | Send messages and interact with Discord servers. | Yes (Bot Token) | | **Telegram** | Send messages via Telegram bots. | Yes (Bot Token) | | **Microsoft Teams** | Send messages and notifications to Teams channels. | Yes | | **Zoom** | Create and manage Zoom meetings. | Yes | | **SMS** | Send text messages via Twilio, Plivo, Vonage, and other providers. | Yes | ## CRM & Sales Tools Tools for customer relationship management, lead tracking, and outreach. | Tool | Description | API Key Required | | ------------------ | --------------------------------------------------------------------- | :--------------: | | **HubSpot** | Full CRM — contacts, deals, companies, and pipelines. | Yes | | **Salesforce** | Enterprise CRM integration. Query and update Salesforce records. | Yes | | **LinkedIn** | Professional networking. Search profiles and company data. | Yes | | **Apollo** | B2B sales intelligence. Find contacts, enrich data, manage sequences. | Yes | | **ActiveCampaign** | Marketing automation and CRM. Manage contacts and campaigns. | Yes | ## Cloud & Infrastructure Tools Tools for cloud services, serverless functions, and automation. | Tool | Description | API Key Required | | -------------- | --------------------------------------------------------------------------- | :-------------------: | | **AWS Lambda** | Invoke serverless functions on AWS. | Yes (AWS Credentials) | | **AWS SES** | Send emails via Amazon Simple Email Service. | Yes (AWS Credentials) | | **Apify** | Web scraping and automation platform. Run pre-built scraping actors. | Yes | | **E2B** | Secure cloud sandboxes for running code. Execute Python, Node.js, and more. | Yes | ## Google Suite Tools Deep integrations with Google Workspace. See individual guides for setup. | Tool | Description | Setup | | -------------------------------------------------- | ----------------------------------------------------------------------------------------- | ----- | | **[Gmail](/docs/tools/gmail)** | Read, search, send, and draft emails. | OAuth | | **[Google Calendar](/docs/tools/google-calendar)** | List, create, update, and delete calendar events. | OAuth | | **[Google Sheets](/docs/tools/google-sheets)** | Read and write spreadsheet data. | OAuth | | **Google Drive** | Search, read, upload, and manage files in Google Drive. Access shared drives and folders. | OAuth | ## Memory Tools Tools for giving agents persistent memory across conversations. | Tool | Description | API Key Required | | ---------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------ | :--------------: | | **Agent Memory** | Built-in memory that persists facts, preferences, and context across sessions. Your agent remembers what users told it in previous conversations. | No | | **Zep Memory** | External memory service with automatic entity extraction, relationship tracking, and temporal awareness. Best for complex, long-running agent relationships. | Yes | ### When to Use Memory Tools * **Agent Memory** — Use when you need simple, built-in persistence (e.g., "remember my name is Alex"). Works out of the box with no external dependencies. * **Zep Memory** — Use when you need advanced memory features like entity graphs, automatic summarization, or memory search across thousands of past interactions. ## Coding Tools Tools for writing, executing, and analyzing code. | Tool | Description | API Key Required | | ---------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------- | :--------------: | | **Coding Tools** | Suite of code generation, execution, and analysis capabilities. Agents can write Python, JavaScript, SQL, and more, then execute it in a sandboxed environment. | No | ### Example: Data Processing Agent ```bash theme={null} # Agent can: # "Write a Python script to clean this CSV data" # "Generate a SQL query to find duplicate records" # "Create a chart from this dataset" ``` ## Orchestration Tools Tools for managing complex agent workflows and parallel execution. | Tool | Description | API Key Required | | ---------------------- | -------------------------------------------------------------------------------------------------------------------------- | :--------------: | | **Parallel Execution** | Run multiple tool calls simultaneously. Speeds up workflows where independent operations can happen at the same time. | No | | **Workflow Tools** | Define multi-step workflows with branching logic, loops, and conditional execution. Useful for complex business processes. | No | ### When to Use Orchestration Tools * **Parallel Execution** — When your agent needs to call multiple APIs or tools at once (e.g., search three databases simultaneously). * **Workflow Tools** — When you need structured, repeatable processes with error handling and retry logic. ## Other Tools | Tool | Description | API Key Required | | ------------------- | --------------------------------------------------------------- | :--------------: | | **Google Maps** | Geocoding, directions, place search, and distance calculations. | Yes | | **OpenWeather** | Current weather, forecasts, and historical weather data. | Yes | | **Calculator** | Mathematical calculations. No setup required. | No | | **File Operations** | Read and write files. Useful for document processing workflows. | No | ## Configuring API Keys Most tools require an API key. You can configure these in two ways: ### Per-Tool Configuration When adding a tool to an agent, enter the API key in the tool's configuration panel. ### Environment Variables Set API keys as environment variables for your deployment: ```bash theme={null} TAVILY_API_KEY=tvly-xxxxxxxxxxxxx FIRECRAWL_API_KEY=fc-xxxxxxxxxxxxx OPENAI_API_KEY=sk-xxxxxxxxxxxxx SERP_API_KEY=xxxxxxxxxxxxx ``` ## Next Steps * **[API & Webhooks](/docs/tools/api-webhooks)** — Connect the agent to any external API, CRM, or webhook: define named endpoints (OpenAPI import), fire a webhook, or make free-form requests. * **[Custom Tools](/docs/tools/custom-tools)** — Connect external tool servers via MCP. * **[Google Calendar](/docs/tools/google-calendar)** — Step-by-step OAuth setup for calendar integration. * **[Gmail](/docs/tools/gmail)** — Connect email capabilities to your agent.