All notable changes to the Dialectical Cognition Framework project.
The format is based on Keep a Changelog.
2.9.2 (2026-02-05)
2.9.1 (2026-01-20)
2.9.0 (2026-01-19)
2.8.0 (2026-01-19)
2.7.0 (2026-01-19)
2.6.0 (2026-01-19)
2.5.0 (2026-01-19)
2.4.4 (2026-01-19)
2.4.3 (2026-01-19)
- update settings.example.json and link checker (#44) (ec2e7de)
- use Zenodo concept DOI in CITATION.cff (#46) (2797591)
2.4.2 (2026-01-19)
2.4.1 (2026-01-19)
2.4.0 (2026-01-19)
2.3.2 (2026-01-19)
- remove docs/ references (gitignored, not visible on GitHub) (#33) (6790910)
- remove github-release.md skill (belongs in github-repository-setup) (#30) (ecbef21)
- repair closing code fences across all markdown files (#34) (6c063d5)
- upload PDF before CITATION.cff update in release workflow (#35) (70808dd)
2.3.1 (2026-01-19)
2.3.0 (2026-01-19)
- add final polish workflows and configurations (#13) (de8cd10)
- add Release Please for automated releases (#12) (2453160)
- improve EPUB generation and validation (#22) (5e583ab)
- make cover image and bibliography optional (#18) (2630067)
- resolve Amazon KDP workflow parsing issue (#17) (4482b85)
- resolve workflow failures (#24) (e6ebc55)
No unreleased changes
Reorganized 21 /dcf skill modes from flat list into 5 intuitive categories:
| Category | Modes | Purpose |
|---|---|---|
| Evaluation & Review | review, checkpoint, self-review, refine |
Quality gates before proceeding |
| Problem Solving | debug, unstick, simplify, diagnose |
Getting past blocks |
| Design & Analysis | architect, tradeoffs, assumptions, premortem, challenge, decide |
Making good decisions |
| Learning & Exploration | learn, onboard, explain |
Building understanding |
| Session Management | compact, context-health, retro, skill |
Context hygiene and capture |
- Added category headers and descriptions to
/dcfskill - Updated README.md mode listing to match new structure
- Updated DCF_ESSENTIALS.md mode listing to match new structure
Following audit against Paul-Elder framework for Socratic questioning:
- Evidence - "How do you know? What evidence supports this?" (probing reasons and factual foundations)
- Consequences - "What are the consequences? What if you're wrong?" (tracing implications and outcomes)
- Updated Socratic Toolkit in
/dcfskill from 5 to 7 operations - Added Meta-Question to treatise table (was in glossary only)
- Added glossary entries for Evidence and Consequences in both DCF_GLOSSARY.md and treatise
- Added Paul & Elder citations to references.bib (2006 book, 2007 journal article)
- Added "Classical and Modern Synthesis" subsection explaining DCF's 7-operation approach
- Provides academic grounding for combining classical Socratic operations with modern pedagogical framework
Three major philosophical frameworks integrated to strengthen DCF's theoretical grounding:
- Popper's Critical Rationalism - Falsificationism grounds DCF's challenge operations (try to disprove, not confirm)
- Argyris's Double-Loop Learning - Explains why metacognition matters (question the frame, not just fix within it)
- Gadamer's Hermeneutics - Fusion of horizons explains why dialogue changes both parties (partnership, not extraction)
New treatise sections:
- "The Hermeneutic Foundation" in Thinking Mirror chapter
- "The Critical Rationalist Foundation" in Socratic Method chapter
- "Single-Loop and Double-Loop Learning" in Metacognition chapter
New glossary entries: Critical Rationalism, Double-Loop Learning, Falsificationism, Fusion of Horizons
New citations: Popper (1959, 1963), Argyris (1977), Argyris & Schön (1978), Gadamer (1960)
- PDF now 193 pages (up from 190)
- Socratic operations expanded from 5 to 7 (now complete per Paul-Elder framework)
- Bibliography now includes 25 sources (up from 18)
- Elevated "What question should I be asking?" to a core DCF principle
- Added to /dcf skill as 5th core principle and Socratic Toolkit operation
- Added dedicated "The Question" section to skill response format
- Added Meta-Question entry to DCF_GLOSSARY.md
- Added as 6th principle in DCF_ESSENTIALS.md
- Added as 8th principle (Metacognitive) in treatise with glossary entry
/dcf refine- Iterate deliberately on output without drift or infinite loops/dcf self-review- Have Claude review its own output (activates evaluation mode)
dcf-workflow- Shell script for chaining modes with checkpoints.claude/scripts/README.md- Script documentation.claude/settings.example.json- Example hooks configuration
- Converted from procedural (scripted steps) to principle-based (outcomes)
- Reduced from 22 modes to 21 modes (more focused)
- Reorganized into 4 categories (was 6):
- Core modes (13)
- Design & Analysis (3)
- Learning & Onboarding (2)
- Session Management (4)
- Archived procedural version to
.claude/skills/archive/dcf-procedural.md
new-project: onboard → architect → premortem (was learn → premortem → review)debug: debug → assumptions → simplify (was debug → simplify → decide)decision: tradeoffs → challenge → decide (was review → challenge → decide)session-end: context-health → compact → retro (was just retro)
- Updated skill description to reflect principle-based approach
- Updated principle count to 16 (was 15)
- Renumbered Collaboration and Agentic Era principles
/dcf diagnose- Identify which anti-pattern you're falling into/dcf onboard- Guided exploration of unfamiliar codebase/domain/dcf decide- Reach closure when you have options/dcf assumptions- Deep excavation of hidden assumptions/dcf simplify- Reduce complexity to essential/dcf context-health- Assess and address context rot- Workflow Composition - Recommended skill sequences for common scenarios
- See Also hints - Each skill now suggests related skills that chain well
- Updated skill reference to reflect 22 modes across 6 categories
- Added Workflow Composition to glossary with index entry
- Anti-Pattern 13: Context Rot - Conversation context degrading over time
- Anti-Pattern 14: Knowledge Gatekeeping - Tribal knowledge blocking AI effectiveness
- DCF_VISUAL_GUIDE.md - New Skill Workflows diagram with 8 recommended sequences
- DCF_ESSENTIALS.md - Skill Workflows quick reference table
- DCF_GLOSSARY.md - New terms: Learning Accelerator, Context Rot, Knowledge Gatekeeping, Workflow Composition, Trust-Speed-Rigor Triangle, Adoption Stages, Shared Context Infrastructure, Test Coverage as Scaffolding
- DCF_TEAM_ADOPTION_PLAYBOOK.md - Core adoption principles: Demonstration Over Explanation, Adoption Stages, Trust-Speed-Rigor Triangle
- DCF_FAQ.md - Workflow composition guidance
- PDF now 190 pages (up from 180)
- Skill modes expanded from 16 to 22
- Anti-patterns expanded from 12 to 14
- Prompt Reference Appendix (NEW) - 8 tables of essential prompts organized by situation
- Clarification, Challenge, Plan Review, Learning, Debugging, Code Review, Synthesis, Meta
- Makes the treatise more self-contained as a practical resource
- Framework Comparison Tables - 3 new tables in Part X
- Comparison matrix across dimensions (DCF vs Prompt Eng vs Ralph vs Plan Mode)
- Situation-based framework selection guide
- Anti-patterns and risks by framework
- Exercise 8: The Silence Test - Tests what practitioners have internalized vs. depend on AI for
- Operationalizes the scaffolding principle
- Includes tracking sheet and reflection questions
- Failure Modes Chapter - Expanded from 4 brief modes to 12 detailed anti-patterns (~350 lines)
- Each pattern includes: What It Looks Like, Symptoms, Root Causes, The Fix
- New patterns: Socratic Theater, Mirror Narcissism, Reactive Evaluation, Infinite Refinement, Rubber Stamping, Complexity Creep, Goal Drift, Abstraction Addiction, Reinvention Addiction
- Quick Reference detection table
- Meta-Fix section on self-awareness
- Anticipatory Calibration Section - Metacognitive skill of forming expectations before prompting
- Session Continuity Section - Pre-compaction documentation strategy
- Pattern-to-Skill Section - Capturing effective patterns as reusable skills
- Agent Ecosystem Chapter - Specialized agents (Explore, Plan, code-reviewer, code-architect)
- When to use which agent
- Background agents: sync vs async execution
- Multi-agent orchestration patterns
- Configuration Philosophy Expansion
- Hooks: automated DCF triggers
- Settings configuration (.claude/settings.json)
- Model selection (haiku/sonnet/opus) as DCF decision
- Session lifecycle and context hygiene
- IDE integration patterns (VS Code)
- Glossary Additions - 5 new terms: Background Agents, Hooks, Pattern Capture, Reinvention Addiction, Session Lifecycle
/dcf refine- Iterate on output that needs improvement/dcf retro- End-of-session reflection and learning capture/dcf compact- Prepare for session compaction/dcf skill- Capture an effective pattern as a reusable skill
- DCF_FAQ.md - New "Claude Code Integration" section with /dcf skill documentation
- DCF_VISUAL_GUIDE.md - New diagrams: /dcf skill modes, agent selection guide, model selection
- DCF_FRAMEWORK_INTEGRATION.md - New sections: Agent Ecosystem, Hooks, Model Selection, Session Lifecycle
- DCF_ANTI_PATTERNS.md - Added Anti-Pattern 12: Reinvention Addiction
- DCF_GLOSSARY.md - New terms: Pattern Capture, Reinvention Addiction, Hooks, Session Lifecycle, Background Agents, Anticipatory Calibration
- DCF_PROMPT_LIBRARY.md - Session Continuity Prompts, Skill Creation Prompts sections
- DCF_ESSENTIALS.md - Agent selection table, model selection table, skill creation guidance
- Examples Directory - Completed case studies demonstrating DCF in action
- Debugging Race Condition (Software Development)
- API Architecture Decision (System Design)
- Learning Rust Ownership (Learning New Technology)
- DCF Research Agenda - Open questions and research priorities
- DCF Video/Podcast Script - Ready-to-use script for media production
- CONTRIBUTING.md - Contribution guidelines for the community
- CHANGELOG.md - This file
- PDF now 180 pages (up from 150)
- Index now contains 64 entries (up from 56)
- Anti-patterns expanded from 4 to 12
- Exercises expanded from 7 to 8
- Completed DCF_CASE_STUDY_TEMPLATE.md (was truncated)
- THE_ARCHITECTURE_OF_THOUGHT.tex - Full treatise on the Dialectical Cognition Framework
- Part I: Foundations - Extended mind thesis, cognitive partnership
- Part II: The Socratic Core - Elenchus, maieutics, dialectic, aporia
- Part III: Recursive Refinement - Iteration methodology
- Part IV: Agentic Adaptations - Checkpoints, trust calibration
- Part V: Theoretical Integration - Multi-level framework analysis
- Part VI: Measurement - Synergy Index, metrics
- Part VII: Applications - Domain-specific guidance
- Part VIII: Future Directions - Research agenda
- Initial README with project overview
- MIT License
This project uses Semantic Versioning:
- Major version: Significant changes to core DCF principles
- Minor version: New resources, extensions, substantial improvements
- Patch version: Bug fixes, typos, clarifications
- Translation to other languages
- Interactive web-based tools
- Video tutorials
- Community-contributed case studies
- Empirical validation studies
- Integration with additional AI tools
None yet. Open an issue to suggest features!
See CONTRIBUTING.md for how to contribute to this project.
From the Dialectical Cognition Framework (DCF) https://github.com/domelic/architecture-of-thought