Laboratory work on JavaScript

Project Concept A Telegram Mini App needs to be developed in the RPG genre, where users can: level up their character; complete quests; collect rare artifacts and equipment; open treasure chests for rewards; improve attributes; participate in seasonal events and rankings. The app should be implemented as a Telegram Mini App with a mobile interface and scalable backend. Technical Requirements A Telegram Mini App in the RPG genre needs to be implemented with the following functionality: authorization via Telegram; character creation and development; leveling system and attribute upgrades; mission and quest completion; inventory and equipment; system of artifacts of varying rarity; opening reward chests; in-game economy and currency; player rankings; administrative panel for content management; adaptive interface for mobile devices; integration with Telegram Bot API. Required Experience at least 2 years of commercial development; Go and React (TypeScript); Telegram Mini Apps and Telegram Bot API; PostgreSQL; Docker, Linux, CI/CD; integration of third-party APIs; experience in developing production solutions; use of AI tools for development (Cursor, Claude Code, Codex, etc.). Preferred Qualifications experience with high-load systems; integration of payment systems; integration of AI / LLM services; experience with analytics and monitoring. Important Along with your application, please send relevant cases of Telegram Mini Apps or Telegram Bot projects. Applications without a portfolio or examples of completed work will not be considered. The final decision on the selection of the contractor will be made primarily based on your cases and experience in implementing similar projects.

This is an unfinished plugin for After Effects: https://drive.google.com/drive/u/0/folders/1Nq9a672OK6ep9Com1lqelUhZxbx3_3f1 If I understand correctly, this is a CEP plugin (Adobe Extension) for After Effects. It is an HTML/JavaScript panel through Adobe CSXS that interacts with After Effects via ExtendScript. The goal was to automate the video creation process in the "maze challenge" style, reference on YouTube GOALRUSH-f8x or similar. Manually edited video: https://drive.google.com/file/d/1Ylzvax6w2JGDwaBqeCdbD49jh5wB5buY/view?usp=sharing File of one of the projects: https://drive.google.com/file/d/15vKd1L9VfHtDqK2wtLQdYEmDpH1d8Jvy/view?usp=sharing Besides the "bugginess" of the system due to vibe coding (so vibe coders are out), such details did not suit the device (real edits): https://docs.google.com/spreadsheets/d/18Svc6GoQ34PgLe1HY7gNX5tPLH0Cpv9E4mfjCRBCMFU/edit?gid=0#gid=0 It is necessary to create a system that will generate projects in After Effects with all the necessary video components listed in Base.pdf, more detailed logic in the form of photos, text documents, and mind maps for understanding all possible variations on the disk: https://drive.google.com/drive/folders/1Uf4Pnw3SKmwpmhL7sjPZFnfAbN0arUjr?usp=sharing Even more detailed (with a more in-depth descriptive part and links to time codes in references) the technical specifications will be finalized in the project workspace The final result is a system that creates videos: from 40 seconds to 1 minute, completely unique content for each creation, without AI metadata in the video file, it should have a clear plot (traps, dead ends, various emotions of celebrity characters as a result of some plot twists in the video), sound design according to the plot (screams and other emotions, explosions, etc.), stylistically not differing from the reference video from the channel.

Description: It is necessary to create a JavaScript script for the Tampermonkey extension. The script will work in an internal CRM system.Logic of operation: The script must read the unique text ID of the current active dialogue on the page (there are a total of 5–10 different IDs). Depending on the ID, the script takes the corresponding text instruction (System Prompt) from the settings. The mapping of parameters by ID should be moved to a separate convenient settings window for the script. When a new message appears in the chat window, the script sends this text along with the prompt via the OpenAI API (model gpt-4o-mini). The received response is inserted into the text input field and initiates sending with a random delay of 20–45 seconds to simulate the natural operation of the operator.Work is purely with text. Budget — 6000 UAH. I am waiting for proposals from developers with experience in working with the OpenAI API and writing browser automation scripts.