Skip to content
Uvilo OS
published Visibility public Owner erik@uvilo.com Approver _ Created _ Updated _

Forge Charter

What Forge Is

Forge is an agentic system for building, automating, and running organizations. It is the single source of truth for an organization’s vision, mission, values, strategic plans, and operating procedures — and it contains the agents to build, test, execute, and evolve them.

Uvilo OS is the first instance of Forge, designed to build, automate, and run Uvilo Inc. and its flagship product, Uvilo AI. The Forge layer itself is general-purpose; Uvilo-specific information lives in department folders that plug into the Forge framework.

Values

Human sovereignty. Humans make the decisions that matter; agents execute. The system exists to amplify human judgment, not replace it.

Clarity. Decisions are traceable from research through specification to outcome.

Stewardship. Consider the needs of all stakeholders — customers, personnel, management, shareholders, and community.

Continuous improvement. Processes are never finished. The system researches, implements, and improves its own operations on an ongoing cycle.

Accessibility. Power without complexity — the system’s sophistication should serve its users, not burden them.

Goals

  • Serve as the single source of truth for all organizational information
  • Structure the organization into departments with cross-functional links
  • Automate the research, implementation, and continuous improvement of organizational processes
  • Run as many organizational functions as possible
  • Provide an easy, repeatable process to plan, create, and manage new features and initiatives
  • Optimize the performance and success of the organization

Scope

Forge is an orchestration and partial execution layer. It coordinates agents, knowledge, and workflows across the organization. It is not a replacement for specialized external systems — analytics, accounting, payroll, and other well-defined domains with mature tools should be integrated, not rebuilt.

Forge handles specification, planning, and verification. Specialized execution (code editors, compliance platforms, financial systems) is delegated to external tools, and Forge checks their outcomes.

Forge must be able to publish outward — internal documentation for personnel, customer-facing information and help sites for the public. What is published and to whom is defined at the organization level.

The system is designed for smaller, AI-first organizations like SMBs where the boundary between human and agent work is actively evolving. As AI systems get smarter, the balance of human involvement will shift — Forge should accommodate that evolution rather than lock in today’s assumptions.

Requirements

  • Provider independence. Every component — LLM, chat interface, documentation engine, deployment platform — must be replaceable. The system defines interfaces, not implementations. Build on open standards and open source so components can be swapped without rearchitecting.
  • Cost efficiency. Token usage and infrastructure costs must be measured and optimized. Agents should carry only the context they need.
  • Tool integration. Forge must connect with all platforms and tools the organization uses today and in the future, bringing data into a single place for analysis and decision-making.
  • Multi-user. Multiple users should be able to use and contribute to the system simultaneously with automatic merging, version control, and archival.
  • Accountability. Every project has an owner and an approver. Those two people are ultimately responsible for the project’s direction and outcomes.
  • Transparency. Every decision is traceable. Research documents show how conclusions were reached; specs show only the final decisions. Both are preserved.
  • Resilience. The system must recover from agent errors and tool failures gracefully. State persists across sessions; work is never lost to a crashed conversation.
  • Evolvability. Forge must have a meta-process for improving itself — its structure, its adherence to its own requirements, and its adaptation as AI capabilities advance.

Constraints

  • Security is non-negotiable. How sensitive data is handled and what agents can access is defined at the organization level, not by Forge itself.
  • Privacy serves the mission. Privacy is important but must not impede agents from doing their work. What PII is pulled in is an organizational decision, not a system default.
  • Simplicity of use, not of system. The system may be complex; the experience of using it must not be.
  • Inspectability by humans. Personnel must be able to view the skills and other internals of agents in a human-friendly way.