Игорь Зуевич

The speaker is comparing AI coding assistants — Claude Code, Gemini, and Codex — by testing their ability to create image-generation agents. Each tool is being evaluated on how well it writes prompts and structures sub-agents for text-to-image and image-to-image tasks.

The transcript explains the 'Compact' command in AI agent sessions, which compresses the entire conversation history to reduce context window usage without switching sessions. It also covers related commands for full history clearing and automatic triggers for compact execution. The speaker recommends configuring automatic notifications or auto-execution when the context window reaches a certain threshold.

The video demonstrates setting up a Telegram HR bot using Claude Code, walking through the process of creating a bot via BotFather, obtaining a Telegram user ID, and letting the code automatically generate the HR agent. The presenter shows how the system initializes itself by installing dependencies and launching the bot with minimal manual configuration.

The video explains how to create 'living' or 'super' skills for AI agents like Claude Code, Gemini, and Codex that self-update based on user feedback. The presenter demonstrates how these skills collect feedback after each task, store it in log files, and periodically incorporate that feedback into the skill's core instructions. This approach reduces token consumption and ensures skills continuously improve over time.

The speaker describes their use of a scheduling/routing tool called Routins, which they compare to N8N, to automate daily delivery of curated YouTube videos on topics like Gemini and Claude. The system runs autonomously, collecting videos and generating concise summaries of their content without manual intervention.

The transcript explains the Newchat command, which opens a new session while pulling data from previous sessions. It works in conjunction with the checkpoint command, which can be configured to trigger automatically when Newchat is launched. The speaker recommends implementing these commands through skills rather than MD slots for a more professional workflow.

The video demonstrates how to configure AI agents (Gemini, Claude Code, Codex) to work autonomously on scheduled timers without manual interaction. The creator shows a step-by-step process of training agents to perform recurring tasks like content creation and architecture audits, then automating them to run at specific times. Materials and handout files are provided to help viewers replicate this setup.

The video demonstrates setting up Gemini CLI as a free alternative to Claude Code, featuring a 2 million token context window. The presenter walks through logging into an account, configuring autostart, and managing permissions within the CLI tool. The setup process reveals some quirks, including session-specific permissions and an initially empty process list.

The speaker describes a personal knowledge management system using a 'dump' or landfill database where raw ideas, voice messages, and notes are stored. An AI agent then processes and synthesizes these notes weekly to build full-fledged content. The system leverages Telegram as the input interface.

The speaker discusses Andrey Karpathy's post about a major problem in AI agent and RAG system design: the difficulty of searching and structuring knowledge bases. The core issue is that notes and knowledge fragments are stored as disconnected elements, making retrieval dependent on remembering exact keywords used during storage.

The video covers setting up a Gemini CLI agent with multilevel memory and a task scheduler, specifically demonstrating the integration of Telegram with Gemini. The presenter walks through configuring a Telegram bridge by providing API tokens and bot credentials to the agent interactively.

The Convo Life team presents a tool for tracking AI agent token usage and context window status within Telegram bots. The system provides usage thresholds and recommendations for when to compact or reset sessions to avoid exhausting the context window. It can also be visually customized per bot.

The transcript covers the setup of a Codex AI agent with voice message support in Telegram, specifically addressing the limitation of standard ChatGPT subscriptions with Whisper. The speaker walks through obtaining an OpenAI API key to enable Whisper transcription and debugging a file extension error with OGA voice files. By the end, the bot successfully receives and transcribes Telegram voice messages.

The video walks through setting up Gemini CLI, covering tool integrations like Google Workspace and Firecrawl, and explaining how to configure agents for automated task planning. The presenter compares Gemini CLI with Claude Code and explains their reasons for switching. Key advice focuses on customizing tool connections to individual needs rather than copying the presenter's setup.

The video explains the 'Dream Team' command for AI agents, which acts as a memory defragmentation and consolidation tool. It is designed to be run periodically (e.g., weekly) to organize accumulated data and notes, helping the agent 'recover' after extended periods of work.

The speaker explains that ClaudeClaw creates more than just a simple connection between Claude Code and Telegram, but rather builds a comprehensive AI employee system with memory, tool access, and role-based permissions. This system includes specialized sub-agents for different functions like accounting and copywriting, each with appropriate access controls.

The speaker demonstrates creating a Telegram bot by finding the official bot creator, generating an API key, and connecting it to their CloudBot system. The installation process took approximately 16 minutes and provided step-by-step instructions in Russian.

A speaker explains how to train an AI content creation system by manually correcting initial posts and providing feedback to improve output quality. The process involves iterative refinement where the AI learns the user's brand voice and style preferences over time.

The speaker explains their Telegram-based AI bot system that uses existing subscriptions from services like Claude and ChatGPT. They emphasize cost efficiency and clarify that these aren't open-source bots but subscription-based services accessed through Telegram.

The speaker explains how to export content from a Telegram channel using the desktop version of Telegram to gather posts for content analysis. They recommend using the desktop version over mobile apps because it offers more functionality and export capabilities.