loki-mode

About loki-mode

Multi-agent autonomous startup system for Claude Code. Triggers on "Loki Mode". Orchestrates 100+ specialized agents across engineering, QA, DevOps, security, data/ML, business operations, marketing, HR, and customer success. Takes PRD to fully deployed, revenue-generating product with zero human intervention. Features Task tool for subagent dispatch, parallel code review with 3 specialized reviewers, severity-based issue triage, distributed task queue with dead letter handling, automatic deployment to cloud providers, A/B testing, customer feedback loops, incident response, circuit breakers, and self-healing. Handles rate limits via distributed state checkpoints and auto-resume with exponential backoff. Requires --dangerously-skip-permissions flag.

面向 Claude Code 的多智能体自治型初创编排系统。在“Loki 模式”下触发。协调工程、QA、DevOps、安全、数据/ML、业务运营、市场、HR 和客户成功等领域的100+个专业化代理。从 PRD 到完全部署并产生收入的产品，过程中无需人工干预。功能包括用于子代理派发的 Task 工具、由3位专项评审并行进行的代码审查、基于严重度的问题分诊、带死信处理的分布式任务队列、自动部署到云提供商、A/B 测试、客户反馈回路、事件响应、断路器和自愈能力。通过分布式状态检查点和指数退避的自动恢复机制处理速率限制。需要 --dangerously-skip-permissions 标志。