marketplace

Dream Team Flow

AI-powered workflows for your entire team — developers, infra, data, product, QA, marketing, sales. One CLI to install, role-based setup, customizable workflow steps.

Install

# New user — full setup with role selection
dtf install <REPO_URL> --company-config company-config.json

# Existing user — add role and workflow steps
dtf configure

What's Included

Type	Count	Highlights
Agents	29 (8 domains)	Engineering, data, infra, testing, product, marketing, design, operations
Skills	74+	Conventions, performance, code review, security, marketing, official Anthropic (PDF/PPTX/XLSX/DOCX)
Commands	21	Dream Team orchestration, PR review, ticket triage, infra workflows
Scripts	33	Quality gates, terraform plan, memory health, cost tracking
Roles	12	Frontend, Backend, Fullstack, Data, Infra, QA, UAT, PO, Sales, Marketing, Ops

Agents (29)

Organized by domain. Your role determines which load — not all 29.

agents/
├── engineering/     (7)  frontend-dev, backend-dev, architect, pr-reviewer,
│                         api-designer, performance-analyst, migration-planner
├── data/            (4)  data-engineer, data-analyst, pipeline-builder, insights-reporter
├── design/          (2)  ui-designer, ux-researcher
├── infrastructure/  (3)  infra-engineer, ci-cd-engineer, security-auditor
├── marketing/       (4)  marketing-ops, sales-enablement, content-creator, social-strategist
├── operations/      (2)  customer-ops, support-responder
├── product/         (3)  po-analyst, requirements-analyst, sprint-prioritizer
└── testing/         (4)  qa-tester, uat-tester, api-tester, performance-benchmarker

Commands

Command	Purpose
/create-stories	Full lifecycle orchestrator — ticket to PR for one or more tickets
/my-dream-team	Multi-agent team implementation with --lite, --local, --resume flags
/infra-ticket	Terraform workflow — Jira → branch → plan → PR with structured plan summary
/review-pr	Line-level PR review. --full for local builds, --deep for multi-agent
/workspace-launch	Create worktree from Jira ticket + spin up session
/workspace-cleanup	Tear down worktree, tmux, branch
/ticket-scout	Batch sprint triage with story point estimation
/ticket-refine	Deep quality gate for a single ticket
/ticket-examples	Code variation examples from codebase patterns
/tdd	Test-driven development loop (red → green → refactor)
/design-an-interface	3 parallel agents with different constraints ("Design It Twice")
/triage-issue	Bug investigation → root cause → Jira ticket with fix plan
/request-refactor-plan	Interview → codebase exploration → tiny-commit plan → Jira ticket
/evolve	Review tool usage patterns and promote to skills/conventions
/reviewers	Manage PR reviewer assignments per category
/team-stats	Dream Team leaderboard and history
/retro-proposals	Analyze learnings and route improvements
/pr-insights	Surface review patterns from scraped PR data
/scrape-pr-history	Extract structured learnings from merged PRs
/scrape-jira-pushback	Extract learnings from AI ticket reviews
/sync-config	Push config to GitHub (private + sanitized public)

Skills

Conventions & Performance

Skill	Purpose
frontend-conventions	React/TypeScript/Tailwind coding style
frontend-performance	Core Web Vitals, bundle analysis, React rendering, images, caching
backend-conventions	.NET microservices coding style
backend-performance	EF Core queries, N+1 detection, caching, async/await, memory
infra-conventions	Terraform, AWS, WAF, monitoring, ECR, CI/CD patterns
aws-performance	CloudWatch, RDS tuning, auto-scaling, S3 lifecycle, cost optimization
data-conventions	Data engineering, EF Core, pipelines

Workflow Skills

Skill	Purpose
code-insights	Opt-in refactoring nudges + DTO analysis with mermaid diagrams for PRs
code-review-skill	React 19, TypeScript, perf, security, architecture review guides
tdd	Test-driven development (red → green → refactor)
playwright-cli	Browser automation for testing and visual verification
visual-development-workflow	Write code, verify visually, iterate
mermaid-diagram	Diagram creation with syntax validation
memory-hygiene	Review and clean up memory files to keep token costs low