Farna Control
Read-only demoInstall

What is Farna Control?

Farna Control is a self-hosted Delivery Governance Platform for the software development lifecycle (SDLC).

As AI and automation take on more of the work, you need one governed view of how changes move through code, dependencies, security, release, documentation and knowledge, evidence, and remediation. Farna Control records the work and the decisions as they happen across GitHub, Notion, and every other tool in the chain.

You define the pipelines, policies, and review gates. Farna Control handles pipeline execution, credential management, append-only audit logging, and human-in-the-loop (HITL) governance, so the record of who approved what agent change is always complete.

The problem

Software delivery generates huge amounts of AI- and automation-driven work across the lifecycle: triaging issues and dependencies, changing code and configuration, running security and quality checks, preparing releases, maintaining documentation and knowledge, and remediating problems. Issue-to-PR automation is one example, not the whole product. Teams either do this manually (slow, inconsistent) or wire together brittle scripts that skip governance entirely. Either way, when the auditors come asking “who approved this change and what did it touch?”, the answer is scattered across a dozen tools, and reconstructing it is expensive.

The Farna Control approach

Farna Control treats SDLC automation as a governed, auditable lifecycle problem. Lifecycle Maps model the whole process, while pipelines execute the stages that are ready to be managed. Each pipeline step is a discrete, typed, approval-gated agent node. Connectors bridge your tools and write every action into one append-only control record. Model backends route large language model (LLM) calls. HITL checkpoints gate human oversight where it matters, and stamp the approver into the audit trail.

Core promise

You decide what agents can do, where humans need to approve, and what can run autonomously. Farna Control enforces those boundaries and records every decision.

  • Getting Started: install Farna Control and run your first pipeline
  • Pipelines: the directed-graph execution model
  • Lifecycle Maps: how pipelines, external tools, and manual stages fit together across the SDLC
  • Human-in-the-Loop: how review gates govern agent actions
  • MCP: drive Farna Control’s governed surface from Claude Desktop, Cursor, and other Model Context Protocol (MCP) clients
  • Frequently Asked Questions: quick answers to common questions
Next: Examples