Most AI coding tools stop at the editor. Sage starts there and goes all the way to a merged pull request.
Sage is a provider-agnostic delivery automation platform built around a simple insight: the software delivery pipeline — from issue analysis to implementation, quality gates, testing, and PR creation — is a structured, repeatable process. One that AI agents can execute with the same rigor a senior engineer would demand.
The Empty Quadrant
The current landscape splits along two axes: low-code automation on one side, AI coding frameworks on the other. CI/CD lives in its own corner. Nobody has connected them — with delivery semantics, composable agents, and quality gates as first-class citizens.
That is the space Sage occupies.
How It Works
A Run in Sage takes an issue reference and walks it through a configurable pipeline of steps:
- Analysis — understand what is being asked and why
- Planning — architectural decisions, task breakdown
- Implementation — Cursor CLI or other AI runtimes execute the actual code changes
- Quality gates — lint, build, test — with automated fix loops
- Verification — local artifacts produced, CI handoff prepared
- Pull Request — created with full trace of decisions and outcomes
Every step is observable. Every decision is an artifact. Nothing is a black box.
Architecture
Sage is built on Clean Architecture with a Blackboard model at its core — a shared space where specialized agents (analyst, architect, quality guard) contribute typed facts that other agents can read and act on. Each Run has its own isolated blackboard.
The system is fully provider-agnostic via a Ports & Adapters design:
- Work items: Gitea, GitHub, GitLab, Jira
- AI runtime: Cursor CLI, OpenAI, or mock — swappable per run
- SCM: any git host
Configuration Over Code
Workflows are defined as JSON DAGs, not hardcoded pipelines. Config merges in layers — global < project < run override — with versioning, rollback, and a full audit trail. Tech leads stay in control of what the agents are allowed to do.
Who It’s For
- Tech leads and staff engineers who want to automate repetitive delivery work without sacrificing quality standards
- Platform engineers building internal AI delivery tooling for their teams
- Multi-repo startups who need consistency across services
- Teams with compliance requirements — every decision is traced, every artifact is persisted
Status
Active development. Core pipeline is functional; agent composability, the Agent Factory UI, and multi-provider support are in progress.