T-800 Agent — цех Skynet для Cursor: создание команд/skills/субагентов, аудит Cursor и плагинов, автообновление с GitHub.
Agent Cursor
A specification guide for Cursor hooks, which are commands triggered by events such as starting a session or editing a file. It defines the hooks.json format, blocking versus observation behavior, and cloud-safety requirements.
Agent Cursor
A specification guide for installation, validation, audit, and gate scripts written in matching Bash and PowerShell versions. It defines their inputs, outputs, exit codes, and safety rules.
Agent Cursor
Brain librarian for Teams, Dashboard, billing, usage limits, integrations, Bugbot, Security Agents. Use for team plugin design or enterprise features. Readonly authoritative KB answers.
Agent Cursor
Brain librarian for Cursor Agent modes, prompting, Plan/Ask/Debug/Design, Agent Review, checkpoints, Canvas. Use when designing agent-related subagents or commands. Readonly authoritative KB answers.
Agent Cursor
Brain librarian for Cloud Agents, Automations, Hooks, cloud settings. Use when designing cloud workflows, automations, or hook pipelines. Readonly authoritative KB answers.
Agent Cursor
Brain librarian for Rules, Skills, Subagents, MCP, context links. Use when creating or editing rules, skills, subagents, MCP configs. Readonly authoritative KB answers.
Agent Cursor
Brain librarian for Cursor CLI, SDK, Cloud Agent API, programmatic agents. Use when building integrations, CI, or external agent orchestration. Readonly authoritative KB answers.
Agent Cursor
A coordinating agent for creating and managing coding-agent extensions such as skills, commands, subagents, rules, and hooks.
Agent Cursor
Brain librarian for Cursor security, Run Modes, permissions, what not to give agents. Use when designing readonly subagents, hooks gates, or safety rules. Readonly authoritative KB answers.
Agent Cursor
Brain librarian for Agent tools: Terminal, Browser, Search, Canvas. Use when subagents need tool access patterns or MCP browser workflows. Readonly authoritative KB answers.
Agent Cursor
A read-only analyst that inventories an extension's skills, commands, agents, and stored project notes.
Agent Cursor
A coordinator for setting up a shared automation hub and its client. It discovers the project’s memory and target setup, then selects specialists for planning, packaging, and testing.
Agent Cursor
A designer of the internal data format used to send a job from a client to an automation hub. The format describes the request, the command to run, callback details, input data, and references to secrets.
Agent Cursor
A specialized subagent that drafts neutral instructions for Cursor Automations, including instructions for a central hub and a client that builds time-zone settings. Cursor Automations are scheduled or event-driven tasks in Cursor.
Agent Cursor
A smoke-test checklist for checking whether a cloud agent service can receive a request, run it, and send back a result.
Agent Cursor
A coordinator for keeping a knowledge base about Cursor Automations and Cloud Agents current. It collects changes from research, news, scouts, and changelogs, then puts update items into a queue for a maintainer.
Agent Cursor
A read-only design agent for planning a new subagent or another Cursor add-on before files are changed. It chooses the suitable type and defines its name, permissions, model, category, and relationships with other agents.
Agent Cursor
A validation agent for coding-agent subagents and their relationship graph. It checks metadata, unique names, registry entries, links between agents, descriptions, and other delivery requirements.
Agent Cursor
A file builder that turns an approved specification into coding-agent files such as subagents, commands, rules, and skills.
Agent Cursor
An installer that places generated agents, skills, commands, or rules into the correct target plugin or workspace. It first identifies the target type, such as a Cursor workspace, a user-wide Cursor setup, or a generic plugin.
Agent Cursor
A coordinator for building coding-agent add-ons such as subagents, skills, commands, rules, and hooks. It runs a defined pipeline involving research, design, construction, integration, and review.
Agent Cursor
A short question step for unclear research requests. It asks 2–5 questions about models, tools, where results should go, read-only limits, and integrations before deeper work begins.
Agent Cursor
A coordinator for reviewing evidence from previous engineering runs and preparing a structured list of follow-up work. It checks lesson severity with a provided classifier and does not directly edit the queue files.
Agent Cursor
A maintenance assistant for the T-800 agent’s knowledge base, documentation, update queue, and installation checks.
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: