Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
## Sprint 0: Project Setup & Infrastructure (Week 1)**Goal**: Production-ready foundation
### User Stories#### US-0.1: Developer can deploy to production**Priority**: High
**Story Points**: 5
**Description**: Set up CI/CD pipeline so any developer can deploy code to production with a single command.
**Acceptance Criteria**:
- [ ] GitHub Actions workflow configured
- [ ] Deploys on push to `main` branch
- [ ] Automated tests run before deploy
- [ ] Deploy to Vercel (frontend) and Railway (backend)
- [ ] Environment variables configured
- [ ] Can rollback to previous version
**Technical Notes**:
- Use GitHub Actions for CI/CD
- Vercel for frontend (auto-deploy)
- Railway for backend workers
- Store secrets in GitHub Secrets
---
#### US-0.2: Database is provisioned and migrations run**Priority**: High
**Story Points**: 3
**Description**: Set up PostgreSQL database with Prisma ORM and run initial migration.
**Acceptance Criteria**:
- [ ] Supabase project created
- [ ] Database connection string in .env
- [ ] Prisma schema matches blueprint
- [ ] Initial migration applied
- [ ] Seed data loaded
- [ ] Can query database from app
**Technical Notes**:
- Use Supabase free tier
- Prisma for ORM
- RLS policies for multi-tenancy
[... more user stories ...]
Step 2: Structure Data for Target Platform
Linear Format:
{"title":"US-0.1: Developer can deploy to production","description":"Set up CI/CD pipeline so any developer can deploy code to production with a single command.\n\n**Acceptance Criteria**:\n- [ ] GitHub Actions workflow configured\n- [ ] Deploys on push to `main` branch\n- [ ] Automated tests run before deploy\n- [ ] Deploy to Vercel (frontend) and Railway (backend)\n- [ ] Environment variables configured\n- [ ] Can rollback to previous version\n\n**Technical Notes**:\n- Use GitHub Actions for CI/CD\n- Vercel for frontend (auto-deploy)\n- Railway for backend workers\n- Store secrets in GitHub Secrets","priority":1,"estimate":5,"labelIds":["<high-priority-label-id>","<infrastructure-label-id>"],"projectId":"<sprint-0-project-id>","teamId":"<team-id>"}
GitHub Issues Format:
{"title":"US-0.1: Developer can deploy to production","body":"**Priority**: High\n**Story Points**: 5\n\n**Description**: Set up CI/CD pipeline so any developer can deploy code to production with a single command.\n\n**Acceptance Criteria**:\n- [ ] GitHub Actions workflow configured\n- [ ] Deploys on push to `main` branch\n- [ ] Automated tests run before deploy\n- [ ] Deploy to Vercel (frontend) and Railway (backend)\n- [ ] Environment variables configured\n- [ ] Can rollback to previous version\n\n**Technical Notes**:\n- Use GitHub Actions for CI/CD\n- Vercel for frontend (auto-deploy)\n- Railway for backend workers\n- Store secrets in GitHub Secrets","labels":["Sprint 0","Priority: High","Type: Infrastructure"],"milestone":"Sprint 0","assignees":[]}
Jira Format:
{"fields":{"project":{"key":"PROJ"},"summary":"US-0.1: Developer can deploy to production","description":"Set up CI/CD pipeline so any developer can deploy code to production with a single command.\n\nh3. Acceptance Criteria\n* GitHub Actions workflow configured\n* Deploys on push to main branch\n* Automated tests run before deploy\n* Deploy to Vercel (frontend) and Railway (backend)\n* Environment variables configured\n* Can rollback to previous version\n\nh3. Technical Notes\n* Use GitHub Actions for CI/CD\n* Vercel for frontend (auto-deploy)\n* Railway for backend workers\n* Store secrets in GitHub Secrets","issuetype":{"name":"Story"},"priority":{"name":"High"},"customfield_10016":5,"labels":["sprint-0","infrastructure"],"sprint":"<sprint-id>"}}
Step 3: Create Projects/Milestones/Sprints
Before creating issues, set up organizational structure:
Include completed: Yes/No (sync already-completed sprints)
Dry run: Preview changes without creating issues
Examples:
# Sync to Linear with 1-week sprints
/architect:sync-backlog --platform=linear --sprint-duration=1-week
# Sync to GitHub with auto-assignment
/architect:sync-backlog --platform=github --repo=owner/repo --assign=@me
# Dry run (preview only)
/architect:sync-backlog --platform=jira --dry-run
# Export to CSV
/architect:sync-backlog --platform=csv --output=backlog.csv
Platform-Specific Features
Linear
Keyboard Shortcuts Support:
Add keyboard shortcut hints in issue descriptions:
**Quick Actions**:
-`c` - Complete this issue
-`a` - Assign to me
-`l` - Add label
-`p` - Change priority
Cycles Integration:
Map sprints to Linear cycles automatically:
/architect:sync-backlog --platform=linear --use-cycles
# Creates or updates cycles for each sprint
GitHub
GitHub Projects Integration:
Create a GitHub Project board automatically:
Example: "❌ No sprint backlog found. Run /architect:blueprint first."
CSV Export Format
For manual import or tools without API:
Sprint,ID,Title,Description,Priority,Story Points,Acceptance Criteria,Technical Notes,Labels
Sprint 0,US-0.1,Developer can deploy to production,"Set up CI/CD pipeline...",High,5,"- [ ] GitHub Actions workflow configured
- [ ] Deploys on push to main branch","Use GitHub Actions for CI/CD
Vercel for frontend","infrastructure,devops"
Sprint 0,US-0.2,Database is provisioned and migrations run,"Set up PostgreSQL...",High,3,"- [ ] Supabase project created
- [ ] Database connection string in .env","Use Supabase free tier
Prisma for ORM","database,setup"
[...]
Import instructions included in output:
📄 CSV exported to: backlog-export.csv
To import:
- Linear: Import via CSV at linear.app/settings/import
- Jira: Tools → Import → CSV
- GitHub: Use gh-issues-import script
- Asana: Import via CSV at app.asana.com/import
Success Criteria
A successful backlog sync should:
✅ Create all sprints/milestones/projects
✅ Create all user stories as issues/tasks
✅ Preserve story points and priorities
✅ Include acceptance criteria as checklists
✅ Add appropriate labels/tags
✅ Link dependencies (blocking relationships)
✅ Be immediately actionable (team can start work)
✅ Match blueprint structure exactly
✅ Provide links to view in target platform
✅ Handle errors gracefully (partial sync)
Examples
Example 1: Linear (Startup)
/architect:sync-backlog --platform=linear
# Interactive prompts:# - Linear API key: [paste from linear.app/settings/api]# - Team: Engineering (auto-detected)# - Start date: 2026-02-07 (today)# - Sprint duration: 2 weeks# Output: 56 issues created across 8 sprints
Example 2: GitHub Issues (Open Source)
/architect:sync-backlog --platform=github --repo=acme/backend
# Requires: gh CLI authenticated# Creates: Milestones + Issues + Labels# Output: Link to project board
Example 3: Jira (Enterprise)
/architect:sync-backlog --platform=jira
# Interactive prompts:# - Jira URL: acme.atlassian.net# - Email: you@acme.com# - API token: [from id.atlassian.com]# - Project key: PROJ# Output: 56 stories created in PROJ project
Example 4: CSV Export
/architect:sync-backlog --platform=csv --output=backlog.csv
# Output: backlog.csv (ready for manual import)
