Локальная мульти-агентная система на self-hosted LLM (Ollama). 3 канала (Telegram/консоль/MAX), 4 типа моделей (LLM/embed/vision/voice), мультимодальность (текст/PDF/фото/голос), роли Planner/Executor/Critic с рефлексией. Инструменты: почта, Яндекс.Диск, web, OCR, cron-планировщик. Семантическая память (sqlite-vec). Coverage 88%. Без облачных API.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of architecture rules for a layered agent application. The layers separate adapters, core task handling, agents, tools, services, language models, and memory so each part has a clear responsibility.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of project rules for handling input and output operations asynchronously. Asynchronous code lets the application wait for network, file, or database work without blocking other tasks.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of rules for moving repeatable checks and tasks into scripts or continuous integration, which automatically checks code changes. It leaves subjective design decisions to people or AI.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A step-by-step method for finding and fixing software bugs. It requires reproducing the problem, writing a test before the fix, correcting the underlying cause, and checking for regressions.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of documentation rules for keeping project instructions and references consistent. It covers relative file paths, documentation updates, link checks, and the order for describing behavior before changing code.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of rules for handling expected failures in an application, such as service outages, timeouts, or invalid tool results.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of project rules for using Git: how to name branches, write commits, and start and finish tracked tasks.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A checklist and decision process for reviewing a pull request or merge request, which is a proposed code change awaiting approval.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of safeguards for systems that accept user text and can access files or other tools.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of rules for preparing a pull request on GitHub or a merge request on GitLab, which asks for code changes to be reviewed and merged.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A guide for adding or changing assistant skills and prompts in a repository. It explains the required files, mirrored copies and checks that confirm everything stays in sync.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A set of project rules for writing and running automated tests with pytest, Python's testing tool, and asyncio, Python's support for asynchronous code.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A reference guide to the project, its creator, its abilities, its technology, and where its source code is hosted.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A Bash and Linux command reference for macOS and Linux terminals. Bash is a shell used to run commands, combine them into pipelines and automate tasks.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A guide for writing safer Bash scripts for production work. It covers strict error handling, input checks, cleanup, tests with Bats and static checks with ShellCheck.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A code-review checklist for examining code without rewriting it. It checks correctness, security, readability and performance, then labels findings by seriousness.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A read-only assistant for reviewing incoming Yandex or Gmail messages from a chat.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
An email-reply drafting tool that reads a message, considers its context and attachments, and produces a structured reply draft. It does not send the email.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A guide for shortening long text into a summary of three to five main points.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A Russian-language editing guide that makes text sound more natural by removing bureaucratic wording, filler phrases and a machine-like tone.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A mail-summary tool that counts email and groups it by unread status, sender and date. It uses a script to calculate the figures instead of reading every message individually.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A decision guide for choosing between OCR, which reads text from an image, and visual description, which explains what a scene or object shows. It also covers cases where both kinds of understanding are needed.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A scheduler for recurring tasks using cron, a time-based system for running jobs automatically. It can create, list and cancel repeating schedules, but does not support one-time reminders.
radif-ru/ai-multi-agent-system
Skill Claude CodeCodex
A weather lookup guide that uses wttr.in to retrieve current conditions and forecasts for a named city, region or airport code. It covers forecasts for several days, including a week-long view.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: