What is Buildforce?
Buildforce is a context-first framework for Spec-Driven Development that works through CLI initialization and slash commands for AI coding agents.
It materializes user intent into structured specifications, generates implementation plans, and tracks execution with full traceability. By creating a persistent context repository that accumulates important context across sessions, it enables AI agents to learn from past decisions and maintain consistency across features.
Instead of re-explaining architectural decisions every time you start a new feature, Buildforce preserves them within your repository where they're searchable and reusable.
What Makes It Different
| AI Assistant only | AI Assistant + Buildforce |
|---|---|
| Context lost after each session | Context persists in .buildforce/context/ |
| No workflow structure | Flexible but structured workflows |
| Requirements exist only in conversation | User intent captured in spec.yaml with acceptance criteria |
| Plans exist only in specific modes | The captured spec is automatically converted into a plan that the user can iterate on |
| Implementation deviations go untracked | Deviations logged with rationale (Original → Actual → Reason) |
| Architectural decisions forgotten | Decisions preserved and searchable |
| Knowledge lives in individual developer heads | Shared context repository for team-wide knowledge |
| Each feature starts from scratch | Each feature builds on accumulated project context |
Core Concepts
Context Repository
A version-controlled directory (.buildforce/context/) containing YAML files that capture:
- Architectural decisions and rationale
- Design patterns and conventions
- Module documentation and relationships
- Implementation knowledge and lessons learned
This context is searchable via /research and grows with every completed feature.
Spec-Driven Workflow
Instead of vague conversations, Buildforce captures requirements in structured spec.yaml files with:
- Functional requirements - What the feature must do
- Non-functional requirements - Performance, security, scalability needs
- Acceptance criteria - How success will be measured
- Scope boundaries - What's in scope and what's not
Implementation Plans
Auto-generated plan.yaml files define how to build the feature:
- Architecture overview and technical decisions
- File structure (what to create/modify)
- Phased implementation tasks with progress tracking
- Testing guidance and validation criteria
Deviation Tracking
When implementation differs from the original plan, deviations are logged with transparency:
deviations:
- phase: "phase_2"
task: "Create API endpoint"
original: "Use Express.js middleware"
actual: "Used Fastify plugin system"
reason: "Fastify provides better TypeScript support for this use case"
Who Should Use Buildforce?
Buildforce is ideal for:
- Development teams working with AI coding assistants
- Solo developers who want structured workflows and persistent context
- Projects where architectural consistency matters
- Teams that need shared knowledge and documentation
- Anyone building complex features that span multiple sessions