Orchestrated Development Controller

Purpose

The central orchestrator that automatically chains skills based on the type of work being performed. This skill is always active and determines which other skills to invoke and in what order.

When to Use

•Always active by default in this workspace
•
Applies to:
- •New project creation
- •Project migration
- •Feature development
- •Refactoring
- •Bug fixes
- •Pre-release validation

Decision Tree

mermaid

flowchart TD
    START[User Request] --> CLASSIFY{Classify Request}
    
    CLASSIFY -->|New Project| NP[New Project Flow]
    CLASSIFY -->|Migration| MG[Migration Flow]
    CLASSIFY -->|Implementation| IM[Implementation Flow]
    CLASSIFY -->|Spec Change| SC[Spec Change Flow]
    CLASSIFY -->|Pre-Release| PR[Pre-Release Flow]
    
    subgraph NP[New Project Flow]
        NP1[project-vision-normalizer] --> NP2[mvp-scope-guard]
        NP2 --> NP3[spec-driven-planning]
        NP3 --> NP4[assumption-and-risk-hunter]
        NP4 --> NP5[workspace-spec-linter]
        NP5 --> NP6[decision-freeze-governor]
    end
    
    subgraph MG[Migration Flow]
        MG1[system-context-extractor] --> MG2[legacy-behavior-snapshot]
        MG2 --> MG3[behavior-first-analysis]
        MG3 --> MG4[migration-scope-partitioner]
        MG4 --> MG5[spec-driven-planning]
        MG5 --> MG6[workspace-spec-linter]
        MG6 --> MG7[decision-freeze-governor]
    end
    
    subgraph IM[Implementation Flow]
        IM1[implementation-boundary-guard] --> IM2[no-guess-implementation]
        IM2 --> IM3[self-test-generator]
        IM3 --> IM4[autonomous-test-runner]
        IM4 -->|Pass| IM5[Done]
        IM4 -->|Fail: Bug| IM6[self-healing-debugger]
        IM4 -->|Fail: Spec Issue| IM7[spec-violation-detector]
        IM6 --> IM4
        IM7 --> HUMAN{Human Decision}
    end
    
    subgraph SC[Spec Change Flow]
        SC1[spec-auto-updater] --> SC2[documentation-consistency-keeper]
        SC2 --> SC3[project-regression-checklist]
    end
    
    subgraph PR[Pre-Release Flow]
        PR1[regression-and-parity-check] --> PR2[security-and-compliance-baseline]
        PR2 --> PR3[delivery-readiness-gate]
        PR3 -->|GO| RELEASE[Release]
        PR3 -->|NO-GO| BLOCK[Block Release]
    end

Skill Sequences

If Starting a NEW PROJECT

Order	Skill	Purpose
1	`project-vision-normalizer`	Normalize raw ideas
2	`mvp-scope-guard`	Prevent scope creep
3	`spec-driven-planning`	Create formal spec
4	`assumption-and-risk-hunter`	Surface hidden risks
5	`workspace-spec-linter`	Validate spec quality
6	`decision-freeze-governor`	Lock decisions

If MIGRATING a PROJECT

Order	Skill	Purpose
1	`system-context-extractor`	Map existing system
2	`legacy-behavior-snapshot`	Capture current behavior
3	`behavior-first-analysis`	Analyze before changing
4	`migration-scope-partitioner`	Define safe phases
5	`spec-driven-planning`	Spec the new system
6	`workspace-spec-linter`	Validate specs
7	`decision-freeze-governor`	Lock decisions

If IMPLEMENTING a TASK

Order	Skill	Purpose
1	`implementation-boundary-guard`	Stay in scope
2	`no-guess-implementation`	Follow spec exactly
3	`self-test-generator`	Create tests from spec
4	`autonomous-test-runner`	Run and classify tests
5	`self-healing-debugger`	Fix bugs (if needed)
6	`spec-violation-detector`	Detect spec issues (if needed)

If SPEC OR CODE CHANGES

Order	Skill	Purpose
1	`spec-auto-updater`	Update specs (if approved)
2	`documentation-consistency-keeper`	Sync all docs
3	`project-regression-checklist`	Check for regressions

If PRE-RELEASE

Order	Skill	Purpose
1	`regression-and-parity-check`	Validate behavior
2	`security-and-compliance-baseline`	Security audit
3	`delivery-readiness-gate`	GO / NO-GO decision

State Transitions

code

┌─────────────────────────────────────────────────────────────┐
│                    ORCHESTRATION STATE                       │
├─────────────────────────────────────────────────────────────┤
│  current_workflow: [new_project | migration | implementation]│
│  current_skill: [skill-name]                                 │
│  skill_queue: [remaining skills]                             │
│  blocked_on: [human | none]                                  │
│  last_result: [success | failure | needs_input]              │
└─────────────────────────────────────────────────────────────┘

Human Interaction Rules

Humans are only involved if:

Condition	Action
Spec ambiguity is blocking	Request clarification
Spec change approval required	Wait for approval
GO / NO-GO decision needed	Present evidence, await decision
Security/compliance issue detected	Escalate immediately

Default to autonomous execution.

Overrides

The orchestrator can be overridden by:

•User explicit instruction — "Skip spec validation"
•Workspace configuration — Disable specific skills
•Emergency mode — Direct path to hotfix

All overrides are logged and auditable.

Constraints

•Never skip security or compliance skills
•Never proceed without spec on HIGH-risk changes
•Always produce audit trail
•Respect decision freezes

The orchestrator serves the spec, not convenience.