一键导入
optimus-code
optimus-code 收录了来自 cloga 的 12 个 skills,并提供仓库级职业覆盖和站内 skill 详情页。
这个仓库中的 skills
TEMPORARILY DISABLED. Do not use this skill for automatic Optimus fleet dispatch. The Optimus orchestrator passthrough is currently disabled because wait_for_completion can exceed MCP request timeouts; when users ask for optimus-fleet, fleet mode, or automatic orchestration, explain that the skill is disabled and use manual step-by-step execution or direct sub-agents instead.
First-run protocol for the Master Agent. Read this BEFORE using any Optimus MCP tools.
Dispatches a task to a specialized agent role using async-first, non-blocking delegation via the Spartan Swarm.
Integrate Optimus Agent Runtime into external applications via HTTP REST, TypeScript SDK, or CLI contract. Use when a user wants to embed AI agent capabilities into their own app, service, or CI/CD pipeline — without dealing with MCP transport.
Autonomous release gate: runs on a schedule, inspects commits since the last tag, determines whether to release, and executes the release pipeline via the release-process skill.
Generic, config-driven release process for any project type. Covers version bumping, changelog, build, documentation, tagging, and publishing with auto-detection for npm, Python, and Rust projects.
Audits new or modified skills for Optimus compliance before deployment. Use this skill whenever reviewing a skill that has been created or changed — check file paths, MCP tool references, structural quality, and error handling. Also use when doing a quality sweep of all skills in the project.
Orchestrates a parallel Map-Reduce architectural review by spawning multiple specialized expert agents to critique a proposal. Builds on top of the delegate-task skill.
Standard unified VCS workflow — every code change must go through a branch, PR, and merge for traceability.
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Teaches the Master Agent how to build, select, and evolve the agent team. Use whenever you need to create a new role, pick a specialist for delegation, improve an underperforming agent, assign engine/model bindings, or figure out why a delegate_task failed due to weak role context. Also use when the user says "find someone to do X", "we need an expert for Y", or "who can handle Z".
Teaches Master Agent how to inspect swarm runtime state, summarize task health, and flag stale or failed work.