Forge¶

The programmable CI/CD engine for modern teams. Forge is a container-native CI/CD system that uses Go-based logic (via generators) instead of complex YAML templating.

Why Forge?¶

Traditional CI/CD systems rely on static YAML. When you need complex logic (dynamic matrices, monorepo change detection, cross-repo triggers), you end up with thousands of lines of unmaintainable YAML or brittle bash scripts.

Forge changes this by making the pipeline dynamic. You can emit new pipeline steps at runtime using any language (Python, Go, Node), allowing your CI/CD to adapt to your code, not the other way around.

Core Features¶

🚀 Dynamic Pipelines: Emit steps at runtime based on your code's state.
🐳 Matrix Builds: Parallel execution across versions and environments.
🧱 Pipeline Templates: Reusable step definitions to keep pipelines DRY.
🐳 Container Native: Every step runs in a clean Docker container.
🛠️ Live Debugging: SSH/Terminal directly into failing job containers from the Web UI.
🛡️ Org-wide Policies: Automatically inject security scans into every pipeline.
📦 Smart Artifacts: Built-in artifact store with automatic fan-in/fan-out.
📊 Real-time UI: Modern, responsive dashboard with Gantt charts, log streaming, and regression detection.
🩺 Runner Health: Detailed dashboard for monitoring self-hosted agents.
⏰ Scheduled Pipelines: Cron-based triggers for nightly builds and maintenance.

Architecture¶

graph TD
    subgraph IG ["Ingress"]
        CLI[forge CLI]
        UI[Web UI]
        SCM[GitHub/GitLab]
    end

    Scheduler["Scheduler (:8080)"]

    subgraph INF ["Infrastructure"]
        DB[(PostgreSQL)]
        Vault[Vault]
        S3[MinIO/S3]
    end

    subgraph WK ["Workers"]
        Agent1["Agent 1"]
        Agent2["Agent 2"]
    end

    IG --> Scheduler
    Scheduler --> INF

    Agent1 -- "Polls" --> Scheduler
    Agent2 -- "Polls" --> Scheduler

    Agent1 -- "Logs/Status" --> Scheduler
    Agent2 -- "Logs/Status" --> Scheduler

    Agent1 -- "Secrets" --> Vault
    Agent2 -- "Secrets" --> Vault

    Agent1 -- "Artifacts" --> S3
    Agent2 -- "Artifacts" --> S3

The scheduler accepts pipeline submissions, applies org policies, stores jobs in PostgreSQL, and serves the Web UI. Agents poll the scheduler, lease jobs, execute them in Docker containers, stream logs back in real time, and upload/download artifacts to S3-compatible storage. All traffic, including live debug terminals, is routed through the scheduler.