Leia

This tool searches your saved documents' content using a combination of keywords and content to find the most relevant results.

This tool uses Perplexity AI Sonar to perform comprehensive web searches with synthesized, researched answers. Unlike simple search engines, Perplexity provides cited sources and contextual understanding. Best for research questions, current events, fact-checking, and queries requiring authoritative sources. Responses include citations and contextual analysis.

This tool saves a new document to your library, including a generated title and markdown content, for easy retrieval later.

This tool fetches a single document from your library using its ID, returning the title, description, content, and metadata for direct access to specific documents.

This tool edits an existing document in your library. Supports full content replacement, targeted section replacement (by heading), appending content, and find/replace operations. Changes are saved immediately and the document is updated in real-time.

This tool allows the AI to create new folders or update existing folder properties including name, description, and instructions. Folder instructions guide AI behavior for all conversations within that folder. Instructions are screened for safety.

This tool retrieves conversations by ID and provides flexible access to conversation data. You can get the full conversation history, generate a summary, or ask specific questions about the conversation content.

This tool searches your conversation history using full-text search. You can search by conversation title/description, message content, or both. Returns matching conversations with metadata and message snippets. Use get_conversation with a conversation nanoid to retrieve more information.

This tool allows your TeamTeacher agent to make a request to another TeamTeacher agent.

This tool delegates research and search tasks to a specialized subagent with access to multiple search tools (OpenAlex, OpenLibrary, Web Search, Perplexity). Useful for complex research queries that benefit from using multiple search sources or when you want a specialized research assistant to handle the search comprehensively. The subagent can orchestrate multiple tool calls and synthesize findings across academic, web, and book searches.

This tool delegates curriculum and educational standards queries to a specialized subagent with access to multiple educational knowledge bases (IB Diploma Programme (DP), IB Middle Years Programme (MYP), Common Core State Standards (CCSS), Universal Design for Learning (UDL), NGSS, WIDA, and state standards). Useful for cross-curriculum comparisons, comprehensive curriculum research, or when you need deep expertise across multiple educational frameworks. The subagent searches the knowledge bases systematically and returns a synthesized answer plus cited sources — with verbatim quotes for open knowledge bases — so the calling assistant can cross-check and refer back to the underlying material. Set "priority" to trade speed for context window: "speed" (default) runs a fast model; "depth" uses a larger context window and stronger synthesis for big multi-framework comparisons (escalates the model for subscribers).

This tool computes math with real math engines instead of relying on AI prediction — the same "never trust mental arithmetic" approach used by dedicated math tutors. It evaluates arithmetic, fractions, unit conversions, matrices, and statistics; simplifies and expands algebraic expressions; and solves equations symbolically. The AI uses it to verify answer keys, worked solutions, and any numeric claims before presenting them, and results come back as beautifully rendered notation.

This tool searches the web using Tavily, an AI-optimized search engine. It supports multiple search topics (general, news, finance), adjustable search depth, time-based filtering, and can extract full page content. You can filter by specific domains and request AI-generated answers synthesized from results.

Tool to take and edit notes throughout the conversation. This helps the TeamTeacher stay on task and keep track of work done and other notes about the conversation. It acts as a shared scratchpad for you and the TT.

Case summaries of how international criminal tribunals — the ICC, ICTY, ICTR, SCSL, STL, and IRMCT — have admitted, weighed, and excluded digitally derived evidence: video, satellite imagery, intercepted communications, photographs, audio, geolocation analysis, and metadata. Built with the Fenix Project. Each summary captures the evidentiary issues a court grappled with — relevance, probative value, prejudice, authenticity, chain of custody, and corroboration — across procedural stages from Pre-Trial through Appeal. Filter by tribunal, situation, case number, proceedings stage, DDE type, or legal issue; use sparse search to pin exact case numbers like `ICC-01/12-01/15`.

Returns a detailed README for a specific knowledge base tool, including its contents, search capabilities, filtering options, and tips for effective searches. Use this to understand what a knowledge base contains and how to search it effectively before making queries. Supports a "mode" parameter for targeted retrieval — a table of contents, a specific section by heading, or a condensed summary — instead of always reading the full README.

This tool uses image generation models from Google or OpenAI to create images from text descriptions. Supports long descriptions. Choose a size (s, m, l, xl) and an aspect ratio (square, landscape, or portrait); on OpenAI models, small and medium always render as 1024x1024 (so pick large or extra large for non-square output) and large square is also capped at 1024x1024. Generates one to four images per request.