AI DevOps Framework

<!-- SPDX-License-Identifier: MIT -->
<!-- SPDX-FileCopyrightText: 2025-2026 Marcus Quinn -->

# AI DevOps Framework

**[aidevops.sh](https://aidevops.sh)** — An [OpenCode](https://opencode.ai/) plugin and AI operations platform for launching and managing development, business, marketing, and creative projects. 11 primary agents handle automatable work across every domain so your time is preserved for real-world discovery and decisions that AI cannot yet reach.

> **Recommended setup:** [OpenCode](https://opencode.ai/) + OpenAI models. GPT-5.5 is the preferred high-capability model for complex agent work; GPT-5.4 mini is the preferred fast, lower-cost model for triage and routine implementation. Claude models (Anthropic) remain fully supported, and other model providers are evaluated from time to time as their quality, latency, and cost profiles change.

*"Scope a mission to redesign the landing pages — break it into milestones, dispatch workers in parallel, validate each milestone, and track budget across the whole project"*

**One conversation, autonomous project delivery — with enterprise-level security & quality-control.**

Founded by [Marcus Quinn](https://github.com/marcusquinn) on 9th November 2025 to help anyone level-up their AI & Open-Source game.

## **The Philosophy**

**Maximum value for your time and money.** **[aidevops](https://aidevops.sh)** is built on these principles:

- **Autonomous orchestration** - An AI supervisor runs every 2 minutes, dispatching parallel workers, merging PRs, detecting stuck processes, and advancing multi-day missions — no human babysitting required
- **Multi-domain agents** - 11 primary agents covering code, automation, product, SEO, marketing/sales, content/video/social, legal, research, business/accounts, health, and framework operations with 900+ subagents loaded on demand
- **Multi-model safety** - High-stakes operations (force push, production deploy, data migration) are verified by a second cross-provider model before execution — different providers have different failure modes, so correlated hallucinations are rare
- **Resource efficiency** - Cost-aware model routing across OpenAI, Anthropic, Gemini, Cursor, and local models; project-type bundles auto-configure quality gates and model tiers, with budget tracking and burn-rate analysis
- **Self-healing** - When something breaks, diagnose the root cause, create tasks, and fix it. Every error is a live test case for a permanent solution
- **Self-improving** - When patterns of failure or inefficiency emerge, improve the framework itself. Session mining extracts learnings from past sessions automatically
- **Gap awareness** - Every session is an opportunity to identify what's missing — gaps in automation, documentation, coverage, or processes — and create tasks to fill them
- **Git-first workflow** - Protected branches, PR reviews, quality gates before merge. Sane vibe-coding through structure
- **Parallel agents** - Multiple AI sessions running full [Ralph loops](#ralph-loop---iterative-ai-development) on separate branches via [git worktrees](#git-worktrees---parallel-branch-development)
- **Progressive discovery** - `/slash` commands and `@subagent` mentions load knowledge into context only when needed
- **Open-source ready** - Contribute to any project the same way you work on your own. Clone a repo, develop solutions to issues locally, and submit pull requests — the same full-loop workflow works everywhere

The result: an AI operations platform that manages projects across every business domain — absorbing everything automatable so you can focus on what matters.

**Built on proven patterns**: aidevops implements [industry-standard agent design patterns](#agent-design-patterns) - including multi-layer action spaces, context isolation, and iterative execution loops.

## **Why This Framework?**

**Beyond single-task AI.** Most AI coding tools handle one conversation, one repo, one task at a time. aidevops manages your entire operation — dispatching parallel AI agents across multiple repos, routing tasks to domain-specialist agents, and running autonomously for days on multi-milestone projects.

**What makes it different:**

- **Autonomous supervisor** - AI pulse runs every 2 minutes: merges ready PRs, dispatches workers, kills stuck processes, advances missions, triages quality findings — no human in the loop
- **Cross-domain intelligence** - 11 primary agents spanning code, automation, product, business, marketing/sales, legal, content/video/social, research, SEO, health, and framework operations — each with domain expertise and specialist subagents
- **Multi-model safety** - Destructive operations verified by a second AI model from a different provider before execution
- **30+ service integrations** - Hosting, Git platforms, DNS, security, monitoring, deployment, payments, communications
- **Mission orchestration** - Multi-day autonomous projects broken into milestones with validation, budget tracking, and au

... (truncated)