ultra-builder-pro-cli 0.5.0

package/skills/ultra-research/steps/step-40-deployment.md

@@ -0,0 +1,123 @@
+# Step 40: Deployment & Infrastructure
+
+## MANDATORY EXECUTION RULES
+
+- WEB SEARCH REQUIRED — verify infrastructure pricing and capabilities
+- INCLUDE cost estimates with sources
+- DEFINE environments: dev, staging, production
+- WRITE output to spec file BEFORE presenting to user
+- ALL output in English (spec files); conversation in Chinese
+
+## PREREQUISITES
+
+- Step 32 (Building Blocks) completed with [C]
+- `.ultra/specs/architecture.md` §1-6 exist
+
+## SEARCH STRATEGY (MANDATORY)
+
+```
+Search: "{{chosen_hosting}} pricing calculator {{product_domain}}"
+Search: "{{chosen_hosting}} deployment best practices"
+Search: "{{chosen_tech_stack}} CI/CD pipeline setup"
+Search: "{{chosen_tech_stack}} docker deployment production"
+```
+
+## EXECUTION SEQUENCE
+
+### 1. Define Deployment Topology
+
+Map the infrastructure for each environment.
+
+### 2. Define CI/CD Pipeline
+
+Build → Test → Deploy stages with quality gates.
+
+### 3. Estimate Costs
+
+Use web search for real pricing data.
+
+### 4. Write Output
+
+**WRITE IMMEDIATELY** to `.ultra/specs/architecture.md` §7-9:
+
+```markdown
+## §7 Deployment View
+
+### Environments
+
+| Environment | Purpose | Infrastructure | URL Pattern |
+|------------|---------|---------------|-------------|
+| Development | Local dev | [Docker Compose / local] | localhost:X |
+| Staging | Pre-prod testing | [Cloud provider details] | staging.X |
+| Production | Live users | [Cloud provider details] | X.com |
+
+### Production Topology
+- **Compute**: [Service type, instance size, count]
+- **Database**: [Managed service, tier, backup]
+- **Cache**: [Service, tier]
+- **CDN**: [Provider]
+- **DNS**: [Provider]
+
+### CI/CD Pipeline
+
+| Stage | Tool | Trigger | Quality Gate |
+|-------|------|---------|-------------|
+| Build | [Tool] | Push to branch | Compilation success |
+| Unit Test | [Runner] | Post-build | 80%+ coverage, 0 failures |
+| Integration Test | [Runner] | Post-unit | All green |
+| Security Scan | [Tool] | Post-test | No critical/high |
+| Deploy Staging | [Tool] | Merge to main | All gates passed |
+| Deploy Production | [Tool] | Manual approval | Staging verified |
+
+## §8 Crosscutting Concerns
+
+### Logging
+- **Format**: Structured JSON
+- **Fields**: timestamp, level, service, traceId, message, context
+- **Tool**: [Logging framework]
+
+### Authentication
+- **Method**: [JWT / OAuth2 / etc]
+- **Provider**: [Auth service]
+- **Session**: [Strategy]
+
+### Error Handling
+- **Pattern**: Result/Either in domain, global handler in infrastructure
+- **Alerting**: [When and how]
+
+## §9 Cost Estimate
+
+| Service | Monthly Cost | Annual Cost | Source |
+|---------|-------------|------------|--------|
+| Compute | $[X] | $[X] | [URL] |
+| Database | $[X] | $[X] | [URL] |
+| Other | $[X] | $[X] | [URL] |
+| **Total** | **$[X]** | **$[X]** | |
+
+_Assumptions: [user count, traffic, storage]_
+```
+
+### 5. Present to User and Gate
+
+```
+[C] Continue — Deployment defined, proceed to Quality & Risks
+[R] Revise — Adjust infrastructure or costs
+```
+
+**HALT — wait for user response before proceeding.**
+
+### 6. Handle Response
+
+- **[C]**: Load next step: `./step-41-quality-risks.md`
+- **[R]**: Revise, update architecture.md §7-9, re-present
+
+## SUCCESS METRICS
+
+- All environments defined
+- CI/CD pipeline with quality gates
+- Cost estimates with sources
+- Crosscutting concerns (logging, auth, errors) addressed
+
+## NEXT STEP
+
+After user selects [C], read and follow: `./step-41-quality-risks.md`

package/skills/ultra-research/steps/step-41-quality-risks.md

@@ -0,0 +1,135 @@
+# Step 41: Quality Scenarios & Risks
+
+## MANDATORY EXECUTION RULES
+
+- QUALITY SCENARIOS must be specific and testable
+- RISKS need probability AND impact AND mitigation
+- BE HONEST — don't minimize risks to make the plan look good
+- WRITE output to spec file BEFORE presenting to user
+- ALL output in English (spec files); conversation in Chinese
+
+## PREREQUISITES
+
+- Step 40 (Deployment) completed with [C]
+- `.ultra/specs/architecture.md` §1-9 exist
+
+## EXECUTION SEQUENCE
+
+### 1. Define Quality Scenarios
+
+For each quality goal from §1, create testable scenarios:
+
+| Quality Goal | Scenario | Stimulus | Response | Measure |
+|-------------|---------|----------|----------|---------|
+| Performance | Load test | 1000 concurrent users | Response time | p95 < 200ms |
+| Availability | Node failure | One server dies | Auto-recovery | < 30s downtime |
+| Security | Injection | SQL injection attempt | Blocked | 0 success |
+
+### 2. Identify Risks
+
+Categories: Technical, Business, Operational, External.
+
+For each risk:
+- **Probability**: High / Medium / Low
+- **Impact**: High / Medium / Low
+- **Mitigation**: What we'll do to reduce it
+- **Contingency**: What we'll do if it happens anyway
+
+### 3. Identify Technical Debt
+
+What are we knowingly deferring? Document it now so it doesn't get forgotten.
+
+### 4. Write Output
+
+**WRITE IMMEDIATELY** to `.ultra/specs/architecture.md` §10-12:
+
+```markdown
+## §10 Quality Scenarios
+
+| # | Quality Attribute | Scenario | Stimulus | Expected Response | Metric |
+|---|------------------|---------|----------|------------------|--------|
+| Q1 | Performance | [Scenario] | [Stimulus] | [Response] | [Target] |
+| Q2 | Scalability | [Scenario] | [Stimulus] | [Response] | [Target] |
+| Q3 | Security | [Scenario] | [Stimulus] | [Response] | [Target] |
+| Q4 | Availability | [Scenario] | [Stimulus] | [Response] | [Target] |
+| Q5 | Maintainability | [Scenario] | [Stimulus] | [Response] | [Target] |
+
+## §11 Risks & Technical Debt
+
+### Risk Register
+
+| # | Risk | Category | Probability | Impact | Score | Mitigation | Contingency |
+|---|------|----------|------------|--------|-------|-----------|-------------|
+| R1 | [Risk] | Technical | High | High | 9 | [Action] | [Fallback] |
+| R2 | [Risk] | Business | Medium | High | 6 | [Action] | [Fallback] |
+| R3 | [Risk] | External | Low | Medium | 2 | [Action] | [Fallback] |
+
+### Known Technical Debt
+
+| # | Debt Item | Reason for Deferral | Impact if Not Addressed | Address By |
+|---|----------|--------------------|-----------------------|-----------|
+| TD1 | [Item] | [Why deferred] | [Consequence] | [Timeline] |
+| TD2 | [Item] | [Why deferred] | [Consequence] | [Timeline] |
+
+## §12 Architecture Decision Records
+
+| # | Decision | Context | Options | Chosen | Rationale |
+|---|---------|---------|---------|--------|-----------|
+| ADR-1 | [Decision] | [Why needed] | [A, B, C] | [Choice] | [Why] |
+| ADR-2 | [Decision] | [Why needed] | [A, B, C] | [Choice] | [Why] |
+
+### Round 4 Summary
+
+**Quality & Deployment Confidence**:
+- **Deployment readiness**: [X]%
+- **Risk coverage**: [X]%
+- **Quality scenario coverage**: [X]%
+- **Overall R4 confidence**: [X]%
+```
+
+### 5. Write Round 4 Research Report
+
+**WRITE** to `.ultra/docs/research/quality-deployment-{date}.md`:
+
+```markdown
+# Round 4: Quality & Deployment
+
+> **Confidence**: [X]%
+> **Steps completed**: 40-41
+> **Completed**: [date]
+
+## Key Findings
+[Infrastructure decisions + cost summary]
+
+## Top Risks
+[Top 3 risks with scores]
+
+## Technical Debt
+[Items knowingly deferred]
+```
+
+### 6. Present to User and Gate
+
+```
+[C] Continue — Architecture complete, proceed to Research Synthesis
+[R] Revise — Adjust quality scenarios or risks
+```
+
+**HALT — wait for user response before proceeding.**
+
+### 7. Handle Response
+
+- **[C]**: Load next step: `./step-99-synthesis.md`
+- **[R]**: Revise, update architecture.md §10-12, re-present
+
+## SUCCESS METRICS
+
+- Quality scenarios are specific and testable
+- At least 5 risks identified with mitigation
+- Technical debt documented with timelines
+- ADRs capture key architecture decisions
+- Round 4 research report written
+
+## NEXT STEP
+
+After user selects [C], read and follow: `./step-99-synthesis.md`

package/skills/ultra-research/steps/step-99-synthesis.md

@@ -0,0 +1,188 @@
+# Step 99: Research Synthesis & Distillate
+
+## MANDATORY EXECUTION RULES
+
+- VERIFY all spec files are complete — no [NEEDS CLARIFICATION] markers
+- GENERATE distillate — token-efficient summary for /ultra-plan consumption
+- VALIDATE field-level completeness — not just "file exists"
+- THIS IS THE FINAL STEP — quality gate before handoff to /ultra-plan
+- WRITE all outputs BEFORE presenting to user
+- ALL output in English (spec files); conversation in Chinese
+
+## PREREQUISITES
+
+- All selected steps completed with [C]
+- Spec files written: discovery.md, product.md, architecture.md
+
+## CONTEXT BOUNDARIES
+
+- Focus: Verify completeness, generate distillate, produce quality summary
+- This step does NOT add new research — it synthesizes existing output
+- Distillate is the primary artifact consumed by /ultra-plan
+
+## EXECUTION SEQUENCE
+
+### 1. Spec Completeness Validation
+
+Read each spec file and validate field-level completeness:
+
+**discovery.md Checklist**:
+- [ ] §0 Problem Validation — has demand signal + confidence
+- [ ] §1 Opportunity Space — has scored opportunity map
+- [ ] §2 Market Assessment — has TAM/SAM/SOM with sources
+- [ ] §3 Competitive Landscape — has comparison matrix + Porter's
+- [ ] §4 Product Strategy — has vision + trade-offs + defensibility
+- [ ] §5 Assumptions — has prioritized assumptions + experiments
+
+**product.md Checklist**:
+- [ ] §1 Problem Statement — has core problem + impact
+- [ ] §2 Personas — has 2-3 personas with goals + pain points
+- [ ] §3 Scenarios — has 3-5 scenarios with flows
+- [ ] §4 User Stories — has stories with acceptance criteria
+- [ ] §5 Feature Scope — has MVP features + exclusions
+- [ ] §6 Success Metrics — has North Star + business + user metrics
+
+**architecture.md Checklist**:
+- [ ] §1-3 Context — has quality goals + constraints + system context
+- [ ] §4 Solution Strategy — has tech stack with rationale
+- [ ] §5-6 Building Blocks — has modules + runtime scenarios
+- [ ] §7-9 Deployment — has environments + CI/CD + costs
+- [ ] §10-12 Quality — has scenarios + risks + ADRs
+
+**If ANY item fails**: List specific gaps and ask user whether to:
+- [F] Fix — Go back to the relevant step to fill the gap
+- [S] Skip — Accept the gap and proceed (note it in distillate)
+
+### 2. Generate Research Distillate
+
+Create `.ultra/specs/research-distillate.md` — a token-efficient summary optimized for LLM consumption by /ultra-plan:
+
+```markdown
+# Research Distillate
+
+> Generated: [date]
+> Source: /ultra-research steps [list of completed steps]
+> Token budget: <2000 tokens
+
+## Product Core
+- **Problem**: [One sentence — what pain we solve]
+- **For whom**: [Primary persona — one sentence]
+- **Unique angle**: [What makes our approach different — one sentence]
+- **North Star**: [Metric name: target by timeline]
+
+## Scope (MVP)
+### Must-Build
+- [Feature 1]: [One-line description] — traces to [US-XX]
+- [Feature 2]: [One-line description] — traces to [US-XX]
+- [Feature 3]: [One-line description] — traces to [US-XX]
+
+### Explicitly Out
+- [Feature X]: [Why excluded]
+- [Feature Y]: [Why excluded]
+
+## Architecture Decisions
+- **Stack**: [Language] + [Framework] + [Database] + [Hosting]
+- **Pattern**: Functional Core / Imperative Shell
+- **Key trade-off 1**: [Chose X over Y because Z]
+- **Key trade-off 2**: [Chose X over Y because Z]
+
+## Constraints
+- [Technical constraint 1]
+- [Organizational constraint 1]
+- [Regulatory constraint 1]
+
+## Top Risks
+1. [Risk 1]: [Probability] × [Impact] — Mitigation: [Action]
+2. [Risk 2]: [Probability] × [Impact] — Mitigation: [Action]
+3. [Risk 3]: [Probability] × [Impact] — Mitigation: [Action]
+
+## Unresolved Questions
+- [Question 1 — what we still don't know]
+- [Question 2 — what we still don't know]
+
+## Rejected Alternatives
+- [Alternative 1]: [Why rejected — prevents re-proposals]
+- [Alternative 2]: [Why rejected]
+
+## Quality Targets
+| Attribute | Target | Priority |
+|-----------|--------|----------|
+| [Performance] | [p95 < Xms] | 1 |
+| [Availability] | [99.X%] | 2 |
+| [Security] | [Compliance level] | 3 |
+```
+
+### 3. Generate Quality Summary
+
+Display to user:
+
+```
+Research Quality Summary
+========================
+Step 00 (Problem Validation):    ✅ [confidence]%
+Step 01 (Opportunity Discovery): ✅ [confidence]%
+Step 02 (Market Assessment):     ✅ [confidence]%
+Step 03 (Competitive Landscape): ✅ [confidence]%
+Step 04 (Product Strategy):      ✅ [confidence]%
+Step 05 (Assumptions):           ✅ [confidence]%
+Step 10 (User Personas):         ✅ [confidence]%
+Step 11 (User Scenarios):        ✅ [confidence]%
+Step 20 (User Stories):          ✅ [confidence]%
+Step 21 (Features & Scope):      ✅ [confidence]%
+Step 22 (Success Metrics):       ✅ [confidence]%
+Step 30 (Architecture Context):  ✅ [confidence]%
+Step 31 (Solution Strategy):     ✅ [confidence]%
+Step 32 (Building Blocks):       ✅ [confidence]%
+Step 40 (Deployment):            ✅ [confidence]%
+Step 41 (Quality & Risks):       ✅ [confidence]%
+
+Overall: [avg]% confidence
+Spec completeness: [X]/[total] fields validated
+Distillate: .ultra/specs/research-distillate.md
+
+Output files:
+  .ultra/specs/discovery.md
+  .ultra/specs/product.md
+  .ultra/specs/architecture.md
+  .ultra/specs/research-distillate.md
+  .ultra/docs/research/*.md (per-round reports)
+
+Next: /ultra-plan
+```
+
+### 4. Present to User
+
+```
+[D] Done — Research complete, ready for /ultra-plan
+[F] Fix — Go back to fix specific gaps
+[E] Export — Generate additional summary format
+```
+
+**HALT — wait for user response before proceeding.**
+
+### 5. Handle Response
+
+- **[D]**: Research workflow complete. Suggest running `/ultra-plan`.
+- **[F]**: Identify which step to revisit, load that step file
+- **[E]**: Generate requested format
+
+## SUCCESS METRICS
+
+- All spec files validated at field level
+- Research distillate generated (<2000 tokens)
+- Quality summary displayed with per-step confidence
+- No [NEEDS CLARIFICATION] markers remain (or explicitly accepted)
+- All output files verified to exist
+
+## FAILURE MODES
+
+- Skipping validation ("files exist, good enough")
+- Distillate is too verbose (>2000 tokens)
+- Distillate missing rejected alternatives (will cause re-proposals)
+- Not checking for [NEEDS CLARIFICATION] markers
+- Inflating confidence scores
+
+## WORKFLOW COMPLETE
+
+This is the terminal step. After [D], the research workflow is done.
+Next command: `/ultra-plan`