The open list · 77 factories · awesome-software-factories · updated 28 October 2026 · from GitHub
Every software factory.
Systems that turn intent into deployed and maintained software, and the companies that report running one. The list lives on GitHub; this page reads it as it is now. The lights show how dark the ones we have rated run.
Cosmos
Runs event-triggered agent fleets across ticket intake, implementation, verification, deployment, monitoring, remediation, and human approval checkpoints in versioned software-factory workflows.
Factory.ai Software Factory
Connects intake and triage, agent implementation, code and security review, automated QA, release gates, documentation, and incident response across persistent delivery automations.
Felan
Carries team requests, schedules, and system events through contextual planning, implementation, testing, review, pull-request handoff, and operational follow-up using unattended agents in isolated environments.
GitLab Duo Agent Platform
Runs governed agents and repeatable flows across planning, coding, security, merge requests, CI/CD, and deployment within GitLab.
Hoplite
Carries repository tasks through isolated implementation, tests and browser verification, pull requests, review feedback, checks, and merge, with scheduled and webhook-triggered runs.
human
Provides an open-source development rig that carries ideas and defects through specification, implementation, review, CI gates, merge, and deployment.
Last Light
Maintains GitHub repositories through issue triage, pull-request review, health monitoring, and an Architect-Executor-Reviewer feature cycle.
Mastra Factory
Runs persistent coding agents across issue intake, triage, planning, implementation, checks, and pull-request review in configurable repository workspaces.
Sgai
Runs a supervised local software factory where a coordinator delegates a goal to specialist agents and executable success checks determine completion.
SHIP
Runs assigned issues through planning, implementation, CI repair, code review, preview deployment, and acceptance testing to produce verified pull requests with configurable human approval gates.
SuperPlane
Turns backlog work into verified, review-ready pull requests through durable workflow stages that coordinate agents, source control, CI, review, approvals, retries, and feedback.
SWE-AF
Runs goals through product and architecture planning, dependency-aware parallel implementation, QA, review, integration testing, acceptance verification, pull-request creation, and bounded CI repair.
AI-DLC
Runs a harness-neutral software lifecycle with requirements, design, implementation, review, delivery, approval gates, persistent state, and audit evidence.
BMad Method
Guides new and existing software from clarification and right-sized planning through reviewed implementation, correction, and learning.
Conductor
Maintains project context and guides coding agents through approved specifications, plans, implementation, review, and correction.
GSD Core
Runs milestone phases through discussion, research-backed planning, parallel implementation, verification and repair, and pull-request shipping.
Kiro
Turns requirements or bug analysis into design documents, dependency-aware tasks, parallel implementation, and correctness checks across IDE, CLI, and web environments.
Loki Mode
Turns tasks and specifications into delivery contracts, writes sealed checks before implementation, runs bounded repair and specialist review, opens pull requests, and emits signed evidence receipts.
nWave
Guides coding agents through discovery, design, acceptance-test definition, test-driven implementation, and delivery with configurable rigor and human gates.
OpenSpec
Maintains reviewable change specifications from exploration and proposal through implementation, verification, and archival into the durable spec set.
Spec Kit
Structures software work from project principles and feature specifications through implementation plans, tasks, and coding-agent execution.
Spec Kitty
Governs specification-driven software work through planning, isolated worktrees, Kanban state, review gates, and merge controls.
Superpowers
Supplies executable coding-agent workflows for requirements discovery, planning, isolated implementation, testing, debugging, review, and completion.
Appian
Builds governed web and mobile process applications with visual design, data integration, and native deployment controls.
Appsmith
Builds and self-hosts internal applications over databases and APIs, with code-level customization, Git version control, and branch-based deployment.
Base44
Turns conversational requirements into full-stack applications with managed data, workflows, previews, branches, and publishing.
Bubble
Provides a visual no-code environment for building, testing, and hosting full-stack web and native mobile applications.
Budibase
Builds and self-hosts internal tools over connected data, combining visual applications with automations, agents, permissions, and APIs.
FlutterFlow
Visually builds, tests, and deploys Flutter applications for web, iOS, and Android with source-code export.
Layerwise
Turns conversational requirements into full-stack web applications through local code generation, live previews, iterative changes, and managed publishing with built-in database, authentication, analytics, and logs.
Not ratedAsk the show →Lovable
Builds and publishes full-stack web applications through conversational planning, generation, preview, security checks, and managed cloud services.
Mendix
Builds and governs web, mobile, and process applications through model-driven and AI-assisted development.
Microsoft Power Apps
Turns requirements and process models into governed full-stack business applications for web and mobile.
NocoBase
Provides an open-source platform where AI agents and visual builders create relational business systems with workflows and permissions.
OutSystems
Provides a managed low-code lifecycle for designing, testing, deploying, monitoring, and maintaining enterprise applications.
Replit Agent
Builds, tests, fixes, and deploys working applications from conversational requirements in a cloud development workspace.
Superblocks
Takes imported or generated business applications through managed data provisioning, static and agentic security checks, environment promotion, deployment, and continuous dependency monitoring.
v0
Provides a headless app-building API that turns prompts or existing repositories into running sandbox previews, iterates and repairs runtime errors, and deploys applications to Vercel.
Astadia FastTrack Platform
Combines automated code and data conversion, parity testing, and deployment preparation for staged mainframe-to-cloud migrations.
AWS Transform
Coordinates assessment, planning, code transformation, testing, migration, and continuous technical-debt remediation across enterprise workloads.
Blitzy
Produces refactored or migrated code repositories through specification review, multi-agent generation, compilation, runtime validation, and QA.
CodeMie
Connects application assessment, codebase analysis, migration planning, code transformation, test generation, and deployment workflows for legacy modernization.
EvolveWare Intellisys
Extracts and optimizes legacy application metadata, generates technology-neutral specifications and modern Java or C# applications, and validates them through vulnerability, unit, regression, and parallel testing.
GitHub Copilot modernization for Java
Assesses and upgrades Java applications with reusable agent tasks for remediation, build repair, testing, containerization, and Azure deployment.
Google Cloud Mainframe Modernization
Combines mainframe assessment, business-rule extraction, agentic code transformation, behavioral validation, data migration, and Google Cloud deployment.
IBM watsonx Code Assistant for Z
Supports mainframe discovery, documentation, refactoring, COBOL-to-Java transformation, compilation, and semantic-equivalence testing.
Konveyor
Provides an open-source suite for inventorying, analyzing, refactoring, and replatforming applications for Kubernetes and cloud-native environments.
Mechanical Orchard Imogen
Uses source analysis and observed production behavior to generate, validate, and incrementally cut over modern replacements for mainframe workloads.
Moderne
Runs deterministic framework, language, dependency, security, and build migrations across many repositories with reviewed pull-request output.
vFunction
Uses static and runtime architecture context to plan modularization, guide code assistants, generate integration tests, and extract or rewrite Java and .NET services.
Alan
Orchestrates coding agents through ticket intake, sandboxed implementation, CI and QA verification, review, and human-gated merge on a shared delivery context layer.
Not ratedAsk the show →Agent Orchestrator
Coordinates project planning, task delegation, isolated coding-agent workspaces, pull requests, CI, and review feedback through a supervised idea-to-merge workflow.
AgentsMesh
Schedules and supervises coding-agent fleets across isolated worktrees, branches, credentials, and distributed runners.
Archon
Builds repeatable AI-coding harnesses from YAML workflows with planning, implementation, validation, review, and pull-request stages.
Director
Coordinates issue intake, planning, parallel implementation, diff review, and pull-request handoff across isolated Git worktrees on a shared canvas.
Eve Software Factory Template
Coordinates GitHub and Linear work through staged implementation and review stations that produce reviewed draft pull requests.
Fabro
Runs graph-defined coding workflows with branching, loops, model routing, Git checkpoints, verification, sandboxes, and human approval gates.
Factory
Provides a reference issue-to-draft-PR workflow with isolated worktrees, verification, and human merge authority for Claude Code and Codex.
HAR
Adds isolated worktrees, deterministic verification, evidence, and observability to coding-agent workflows through a CLI and MCP interface.
Machinist
Operates coding workflows through durable events, bounded workspaces, retained artifacts, validation, and pull-request handoff.
OrgOps
Bootstraps and coordinates external coding-agent runtimes across pinned hosts and channels with scoped credentials and operator controls.
Overcut
Coordinates agentic software-delivery workflows triggered from Git, ticket, and CI events, with multi-agent orchestration, human approval gates, audit trails, and retrospectives that feed run learnings back into agent memory.
Not ratedAsk the show →Port
Operates an agentic software-delivery lifecycle in which tickets are enriched, scored for AI-readiness, approved by humans, and delegated to coding agents across triage, planning, build, review, and ship stages.
Not ratedAsk the show →Shipfox
Runs event-driven engineering workflows with agent and shell steps, executable gates, bounded feedback loops, isolated runners, and monitoring.
softwarefabrik.io
Runs locally governed software-delivery workflows across planning, controlled agent execution, builds, review, CI correction, approval, merge, policy enforcement, and signed audit evidence.
Superset
Orchestrates parallel coding agents across isolated worktrees with task delegation, scheduled runs, result verification, diff review, and pull-request management.
Tessl
Runs version-controlled agent workflows on schedules and pull requests, with managed execution environments, code-review and CI gates, evaluations, and governed skill distribution.
Warp Factories
Defines cloud software factories as code that move work from intake to mergeable pull requests across configurable agents, models, repositories, approval gates, evaluations, and benchmarks.
CI&T Agentic SDLC
Coordinates business analysis, technical design, implementation, automated testing, validation, and production deployment through a service-led agentic delivery method.
HCLTech AI Force.Software.Mod
Automates legacy-application analysis, code transformation, refactoring, testing, and deployment, with security, compliance, and observability integrated into the delivery process.
Lunatech Legacy Modernization with AI
Defines a seven-phase service method for discovering, rebuilding, parity-testing, migrating, hardening, and handing over legacy-system replacements.
Thoughtworks AI/works
Connects legacy-code analysis, future-state specifications, generated software, evaluations, runtime maintenance, and governance in an expert-led platform.
1,500+ PRs Later: Spotify's Journey with Our Background Coding Agent (Honk, Part 1)
Documents Spotify's repository targeting, agent implementation, formatting, linting, evaluation, review, and production merge workflow, with more than 1,500 merged pull requests and reported migration time savings of 60-90%.
Harness engineering: leveraging Codex in an agent-first world
Reports OpenAI's use of Codex-driven planning, implementation, testing, review, CI repair, merge, deployment, observability, and maintenance to ship a million-line product through roughly 1,500 merged pull requests with no manually written code.
Introducing Murmur
Reports Macroscope's two-month use of cloud-agent fleets for decomposition, implementation, testing, UI verification, pull-request review, CI repair, and merge readiness across 90% of its shipped code.
Minions: Stripe's one-shot, end-to-end coding agents
Documents Stripe's unattended Slack-to-pull-request workflow, including context gathering, implementation, linting, testing, bounded CI repair, and human review, with more than 1,000 fully agent-produced pull requests merged weekly.
Running a Software Factory Efficiently at Uber Scale
Reports Uber's use of managed agents for code review, CI repair, end-to-end pull requests, incident triage, debugging, and maintenance, with measured adoption, quality, and unit-cost outcomes.
No factory matches. Know one that should be here? Add it.
Lights: L0 every room lit, a person at every stage. L5 lights off, nobody reads the code. From the show's graph and the Dark Factory Matrix.
Add a factory
Know one we missed?
The list takes pull requests. Fill this in and we write the entry the way the list wants it, copy it, and open the list on GitHub, where you paste it under its category and propose the change. GitHub makes the fork for you.
The entry
- [Name](https://…) - Description.
Before you send it, read what the list takes: active systems that own a multi-stage software lifecycle, or first-party reports from operating factories. Not single coding assistants or generic agent frameworks.