Skip to content

Strategic Vision

The FCC Agent Team Framework aims to become the standard open-source framework for AI agent orchestration in documentation workflows within three years. This vision is grounded in the belief that documentation is not a secondary concern but a first-class engineering output -- and that AI agents, properly orchestrated, can produce documentation that matches or exceeds human quality at dramatically higher throughput.

Three-Year Horizon

Year 1: Foundation and Adoption (2025-2026)

Status: Largely complete.

Establish the framework's technical credibility and canonical authority over the FCC/R.I.S.C.E.A.R. specifications. Deliver a complete, well-tested, and well-documented package that teams can adopt with confidence.

Key milestones achieved:

  • 147 personas (102 core + 45 vertical) with full R.I.S.C.E.A.R. specifications and 56-dimension profiles
  • 12,100+ tests at 100% coverage
  • Docs-as-code generation producing 1,348 files
  • Three workflow graphs (5, 20, 24 nodes) with BFS and topological traversal
  • Deterministic and AI-powered simulation engines
  • 25 quality gates and 30 capability tags

Remaining Year 1 goal: PyPI distribution (Phase 6) to enable pip install fcc-agent-team without cloning the repository.

Year 2: Integration and Automation (2026-2027)

Expand the framework's reach through plugin architecture, third-party integrations, and real-time collaboration capabilities.

Target outcomes:

  • Plugin system allowing custom simulation engines, template renderers, and governance scorers
  • Integration with CI/CD platforms (GitHub Actions, GitLab CI, Azure DevOps) for automated documentation quality checks
  • Human-in-the-loop simulation where AI-powered personas pause for human review at configurable checkpoints
  • Agent-human feedback loops that improve persona behavior over time

Year 3: Ecosystem and Community (2027-2028)

Build a community around the framework with contributed personas, shared templates, and cross-project orchestration.

Target outcomes:

  • Community persona marketplace where teams publish and share custom personas
  • Cross-project orchestration allowing FCC workflows to span multiple repositories
  • Industry-specific persona packs (healthcare, finance, government, SaaS)
  • Academic research partnerships on AI agent collaboration patterns

Guiding Principles

Open Source

The FCC framework is and will remain open source under the MIT license. Open source is not a marketing strategy -- it is a design principle. The framework's value comes from network effects: more personas, more templates, more workflow patterns, more governance configurations. These network effects require open contribution.

Composable

Every component of the framework is designed to be used independently or in combination. A team that only needs the persona registry can use it without the simulation engine. A team that only needs the quality gates can use them without the docs-as-code generator. Composability reduces adoption friction and supports incremental adoption.

Governance-First

Documentation quality is not optional. The framework embeds governance into every layer: quality gates validate outputs, capability tags classify artifacts, compliance scoring measures adherence, and the Discernment Matrix evaluates judgment quality. Governance is not a bolt-on -- it is foundational.

Agent-Native

The framework is designed for AI agents from the ground up. The 14 consumer-to-agent reinterpretations in the dimension system (e.g., "Demographic Information" becomes "Agent Profile", "Health and Wellness" becomes "Agent Sustainability") reflect a deliberate design choice: these are not human personas repurposed for AI. They are AI agent personas designed with full awareness of what makes AI agents different from human collaborators.

Deterministic by Default

Simulation outputs are reproducible. The mock simulation engine produces identical results given identical inputs. AI-powered simulation adds variability, but traces capture every step for audit and replay. Determinism enables testing, debugging, and compliance verification.

Canonical Authority

This project owns and defines the FCC framework and the R.I.S.C.E.A.R. specification. Downstream projects (including the reference project PAOM) are consumers of these specifications, not authorities. Changes to the canonical specifications originate here and flow downstream.

This ownership model ensures:

  • A single source of truth for specification definitions
  • Controlled evolution of the framework's core contracts
  • Clear contribution pathways for the community
  • Version-tagged specifications that downstream projects can pin to