-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Karthik K Pradeep edited this page Sep 20, 2026
·
2 revisions
OdinRun is a self-hosted, container-native CI/CD runtime built for high visibility, automated failure diagnosis, and intelligent pipeline remediation.
| Section | Description |
|---|---|
| 📖 Getting Started | Installation, Setup Wizard, and Docker Compose modes |
| 📝 Pipeline Schema |
.odinrun.yml and .pipelineos.yml YAML syntax reference |
| 🏗️ Architecture & Design | Core components: API, Runner Fleet, Telemetry, and Adapters |
| 🔗 GitHub Integration | Configuring Webhooks, PATs, and GitHub App credentials |
| 🤖 AI & Auto-Remediation | Setting up Ollama / OpenAI failure diagnosis and auto-retry rules |
| 🛠️ CLI Reference | Usage guide for ./odinrun (start, advanced, stop, doctor) |
| ❓ Troubleshooting & FAQs | Common issues, health checks, and debugging tips |
- Container-Native Execution: Run every stage in isolated Docker containers with live streaming stdout/stderr logs.
-
Dual Pipeline Config Support: Native support for
.odinrun.ymlwith backward compatibility for.pipelineos.yml. - Real-Time Telemetry: Live CPU & Memory utilization metrics pushed directly via WebSockets to a modern React dashboard.
- AI Failure Diagnoses: Automated root cause analysis using local Ollama models or OpenAI.
- Self-Healing Rule Engine: Define automated remediation rules for pattern-matched pipeline failures.
- Flexible Storage Adapters: Works out-of-the-box with SQLite & In-Memory Queues (Simple Mode) or scales to MongoDB & Redis (Advanced Mode).