Skip to content

Best Practices

This section is the rulebook for building, operating, debugging, and scaling the OpenSIN fleet. It is not a loose suggestion shelf. It is the operational memory of the system.

Core Mandates

DocumentFocus
Agent DesignUltimate fleet mandates, no-silo rules, self-healing, test-proof culture
Code QualityExtreme commenting mandate, anti-AI-slop, review discipline
Error HandlingImmediate bug registry, no-assumptions, self-healing escalation
Browser AutomationDevTools-first, anti-bot bypass, Chrome profile law
A2A CommunicationPure agentic paradigm, inbound governance, opencode-only LLM usage

System Reliability & Execution

DocumentFocus
TestingRuntime proof, workflow validation, UI/browser verification
Monitoring & ObservabilityHealth, metrics, evidence retention, alert usefulness
Team OrchestrationParallel vs sequential work, retries, specialist routing
HF Fleet Keep-AliveHugging Face wake strategy, persistence, recovery
CI/CD mit n8n + sin-github-actionZero-billing CI via OCI + n8n

Advanced / Specialized

DocumentFocus
MCP IntegrationMCP transport, safety, integration patterns
Plugin DevelopmentPlugin architecture and extension rules
PerformanceCost, model routing, latency, efficiency
SecuritySecrets, auth boundaries, operator trust
Software 3.0: Neural-BusHigher-level architecture doctrine
SEO PipelineProof-of-work blog publishing pipeline
Simone MCP + PCPM ⭐ PFLICHTSemantic code intelligence + cross-session memory — mandatory for every agent

Before You Call Something “Done”

  • [ ] issue exists and matches the work
  • [ ] code is commented with WHAT / WHY / WHY NOT / CONSEQUENCES
  • [ ] repo-native checks pass
  • [ ] runtime proof exists
  • [ ] screenshots/logs exist where relevant
  • [ ] docs updated if architecture or workflow changed
  • [ ] remaining risk is clearly stated

Final Reminder

The OpenSIN fleet is allowed to move fast only because it is forced to leave evidence, structure, and recoverable knowledge behind. Without that, autonomy becomes chaos.