Changelog¶
All notable changes to the FCC Agent Team Framework, organized by development phase.
v1.0.3 -- Publishing Pipeline Live + For-Teams Expansion¶
Timeline: April 2026
Added¶
- 27 pre-built publication artifacts now committed to
publications/_output/: - 16 book files (4 books × PDF/EPUB/DOCX/HTML)
- Executive package (PDF, DOCX) + presentation (PPTX, HTML)
- Academic package (PDF, EPUB)
- 5 reference cards (PDF via Typst)
.github/workflows/publish.ymlauto-uploads all artifacts to GitHub Releasesdocs/publications/index.md- comprehensive build guide- 4 new
docs/for-teams/files: workflow templates, multi-team governance, scaling guide, shared KB - 2 new sample prompt files: team-coordination, publications-workflow
- 2 new tutorials: building-your-own-book, contributing-to-books
- Publications section in README.md with all download links
- Publishing section in
make helpwith all 6pub-*targets
Fixed¶
- End-to-end publishing pipeline now functional (Quarto 1.5.57 + Typst 0.11.0 + Pandoc 3.2.0)
make helplists all 44 Makefile targets (was missing 11)- Pandoc citation errors resolved via
-f markdown-citationsflag - Pandoc Word/PowerPoint templates generated and committed
Changed¶
- Version 1.0.2 → 1.0.3
.gitignore: publications/_output/ now committed; only intermediates excluded- Individual pub-* targets delegate to build_all.sh for consistency
v1.0.2 -- Multi-Format Publishing & Documentation Overhaul¶
Timeline: April 2026
Fixed¶
- Resolved all 299 mkdocs strict-mode build warnings
- Fixed all persona nav paths (84 entries), 2 guidebook chapter typos, 4 demo nav remaps
- Converted ~183 broken documentation links to GitHub repository URLs
Added¶
- 300+ Mermaid diagrams across all documentation sections (persona interaction maps, category mindmaps, API class diagrams, specification ER diagrams, guidebook/book chapter diagrams)
- Multi-format publishing pipeline (
publications/directory) using Quarto + Pandoc - PDF, EPUB, DOCX output for 3 book series and 24-chapter guidebook
- Executive PowerPoint deck for decision makers
- Academic/educator package (curriculum, workshop guide, student workbook)
- 5 quick-start reference cards (persona categories, CLI commands, workflow patterns, prompt templates, R.I.S.C.E.A.R.)
- 20 persona category overview pages with mindmap diagrams
- MkDocs enhancements: version selector (mike), print-site plugin, dark-mode Mermaid, code annotations, social cards
docs/model-cards/index.md, compliance getting-started guide, NIST RMF crosswalk demo, evaluation overview demorequirements-docs.txt,scripts/fix_mkdocs_nav.py,scripts/fix_doc_links.py,scripts/generate_persona_diagrams.py
Changed¶
- Version 1.0.1 → 1.0.2
- Generated documentation now covers all 102 core personas across 20 categories (previously 24 across 5)
- MkDocs nav includes all 20 core categories with correct file paths
- All external links converted to GitHub URLs
- Makefile extended with
pub-*targets for multi-format publishing
Phase 5.5 -- Dimensions and Documentation Overhaul¶
Timeline: January 2026 - March 2026
Added¶
- 56-dimension persona profiling system across 9 categories (core, behavioral, communication, cultural, decision-making, professional, market, innovative, advanced)
- Dimension profiles for all 24 personas as individual YAML files under
data/personas/dimensions/ - Dimension interpretation guide (
data/personas/dimension_interpretation_guide.yaml) documenting 14 consumer-to-agent reinterpretations - DimensionRegistry class for loading and querying dimension definitions
- PersonaDimensionProfile class for per-persona dimension values
- MkDocs integration for documentation site generation
- Documentation overhaul with comprehensive specification, developer, and roadmap sections
- Strategic architecture review documenting cross-project alignment
Changed¶
PersonaSpecextended withdimension_profilefield- Documentation structure reorganized for MkDocs compatibility
- Test suite expanded to cover dimension system
Phase 5 -- CI/CD, Scenarios, and Documentation¶
Timeline: November 2025 - January 2026
Added¶
- CI/CD pipeline with automated testing, linting, and coverage checks
- 6 starter scenarios (
data/scenarios/starter_scenarios.json) for basic workflow testing - Advanced scenarios (
data/scenarios/advanced_scenarios.json) for complex multi-phase simulations - Scenario models (
src/fcc/scenarios/models.py) with steps, expected outcomes, and validation rules - Scenario loader for JSON-based scenario discovery
- FCCValidator with
from_registry()for registry-aware scenario validation - Docs-as-code generator producing 1,348 documentation files (504 tutorials, 504 prompts, 144 workflows) from 15 Jinja2 templates
- DocGenerator class with per-persona generation, cross-reference pages, and sitemap
- CLI commands:
generate-docs,validate-docs,sitemap - Tutorial and prompt topic definitions (
data/docs/tutorial_topics.yaml,data/docs/prompt_topics.yaml) - Constitution template (
data/docs/constitution_template.yaml) for persona governance - Test suite expanded to 12,100+ tests at 100% coverage
Changed¶
- CLI restructured with Click groups for better command organization
fcc simulatecommand now supports--mock/--no-mockflag- Quality gates expanded to cover all 24 personas (25 gates total)
Phase 4 -- Extended Persona Gallery¶
Timeline: September 2025 - November 2025
Added¶
- 19 additional personas across 4 new categories:
- Integration (7): CIA (Catalog Indexer Architect), UMC (UI Mockup Crafter), STE (Semantic Taxonomy Engineer), TS (Traceability Specialist), BV (Blueprint Validator), RIC (Research Inventory Crafter), GCA (Governance Compliance Auditor)
- Governance (3): DGS (Data Governance Specialist), PTE (Privacy Taxonomy Engineer), AMS (Anti-fact Mitigation Specialist)
- Stakeholder (5): CO (Collaboration Orchestrator), SMC (SAFe Metrics Crafter), EC (Executive Communicator), RS (Roadmap Synchronizer), SCP (Stakeholder Content Publisher)
- Champion (4): RCHM (Research Crafter Champion), BCHM (Blueprint Crafter Champion), UGCH (User Guide Crafter Champion), RBCH (Runbook Crafter Champion)
- Champion architecture with
champion_ofandorchestratesfields on PersonaSpec - Extended workflow graph (
data/workflows/extended_sequence.json) with 20 nodes and 40 edges - Complete workflow graph (
data/workflows/complete_24.json) with 24 nodes and 52 edges including orchestration edges - Cross-reference matrix (
src/fcc/personas/cross_reference.py) with queryable persona-to-persona interactions (upstream, downstream, peers, by_type, between) - Cross-reference YAML (
data/personas/cross_reference.yaml) defining all persona interactions - Cross-reference schema (
data/schemas/cross_reference.schema.json) - Discernment Matrix model (
DiscernmentTrait,RatingDimensions) with 6 traits and 7 rating dimensions - Design Target Factors model (
DesignTargetFactor) with 6 factors and shared rating dimensions - Capability tag registry (
data/governance/tag_registry.yaml) with 30 tags across 7 categories - Tag governance (
src/fcc/governance/tags.py) with three-level taxonomy (capability/category/supercategory) - Quality gates for all 24 personas plus 1 universal gate (25 total)
- Doc context field on PersonaSpec with
primary_action,primary_output,cli_command,example_scenario, andconstitution(hard_stop, mandatory, preferred) - Deliverables and Collaboration structured data on all personas
Changed¶
PersonaSpecextended withcategory,champion_of,orchestrates,doc_context,discernment_matrix,design_target_factors,deliverables,collaborationfieldsRISCEARSpecis now a complete 10-component specification across all personas- Registry supports
by_category()andchampions()queries
Phase 3 -- R.I.S.C.E.A.R. Expansion¶
Timeline: August 2025 - September 2025
Added¶
- 3 new R.I.S.C.E.A.R. components: Role Skills, Role Collaborators, Role Adoption Checklist
- R.I.S.C.E.A.R. completeness gate (QG-RISCEAR-001) validating all 10 components
- Extended R.I.S.C.E.A.R. for all 5 core personas with detailed skills, collaborator descriptions, and adoption checklists
Changed¶
RISCEARSpecdataclass expanded from 7 to 10 fields- Persona YAML files updated with the 3 new components
persona.schema.jsonupdated to include new fields
Phase 2 -- Ownership and Strategic Positioning¶
Timeline: July 2025 - August 2025
Added¶
- MIT license establishing open-source distribution
- CLAUDE.md declaring canonical authority over FCC/R.I.S.C.E.A.R.
- Strategic architecture review documenting relationship with reference project (PAOM)
- Ownership statement in package metadata
Changed¶
- Repository renamed to
l2_fcc_agent_team_ext - Package metadata updated with owner (INFORMATION COLLECTIVE, LLC) and license
Phase 1 -- Foundation¶
Timeline: June 2025 - July 2025
Added¶
- Package structure (
src/fcc/) with proper Python packaging - 5 core personas: RC (Research Crafter), BC (Blueprint Crafter), DE (Documentation Evangelist), RB (Runbook Crafter), UG (User Guide Crafter)
- Base workflow graph (
data/workflows/base_sequence.json) with 5 nodes and 11 edges - WorkflowGraph class with BFS, topological sort, adjacency queries
- PersonaSpec and RISCEARSpec frozen dataclasses
- PersonaRegistry with YAML loading and ID-based lookup
- SimulationEngine with deterministic mock mode
- AI simulation engine with Anthropic and OpenAI support
- AIClient abstraction with mock provider for testing
- Persona-aware prompt generation from R.I.S.C.E.A.R. specs
- Simulation traces with JSON serialization
- Quality gate framework (
src/fcc/governance/quality_gates.py) - CLI tool with
fcc init,add-persona,validate, andsimulatecommands - JSON Schema validation for workflows and personas
- Initial test suite with pytest and pytest-cov
- Makefile with
venv,install-dev,test,linttargets
Related Pages¶
- Roadmap Overview -- phase timeline and future plans
- Release Plan -- upcoming Phase 6-8 details
- Strategic Vision -- where this is all headed