End-to-end autonomous software design, implementation, and review in a single pipeline

Supreme Team

Supreme Team is an AI skill system that drives a complete software lifecycle — design, build, adversarial review, and Azure deployment — through a single orchestrated pipeline. It runs inside AI coding assistants (Claude Code, Cursor, Windsurf, or any tool that loads Markdown skill files) and replaces manual back-and-forth with structured, gated delegation across 35 specialized skills.

Why Supreme Team

Without Supreme Team, asking an AI assistant to "build me an app" produces a single-pass attempt with no structured validation, no adversarial review, and no way to resume if the conversation ends mid-task.

With Supreme Team, the same request flows through a phased pipeline where each deliverable is challenged by adversarial gatekeepers before the next phase begins. Design specs are validated before code is written. Code is security-audited before review. Review findings are evidence-checked before delivery. Every artifact is saved to disk for cross-session resume and audit.

What You Get

• One entry point — tell admiral what you want and it routes through the right phases automatically
• Adversarial quality gates — five gatekeepers challenge every deliverable; approval is earned, never assumed
• Cross-session persistence — pipeline state, deliverables, and audit trails are saved to disk so you can resume where you left off
• Flexible execution — run the full pipeline, any subset of phases, or individual skills in standalone mode
• No platform lock-in — plain Markdown files that work with any AI tool that reads skill definitions
• 14 tech-stack templates — pre-configured backend and frontend stacks from Go/Gin to React/Next.js to Rust/Axum

Skills and Orchestrators

Supreme Team contains 35 skills organized into four sub-pipelines, each managed by its own orchestrator, with admiral as the top-level entry point.

Sub-Pipeline	Orchestrator	Skills	Purpose
Admiral Layer	admiral	2	Top-level orchestration + cross-pipeline validation
Design	commander	7 + 14 templates	Requirements, architecture, API contracts, stack selection
Build	build-management	8	Implementation, testing, security audit, debugging, health checks
Review	code-chief	10	Bug detection, code quality, security, penetration testing, frontend audit, visual QA, DX audit
Azure	azure-provisioner	7	Infrastructure design, deployment, configuration, verification
Session Memory	—	1	Cross-session state checkpoints and accumulated learnings

Every sub-pipeline orchestrator delegates to its specialists in sequence and validates each phase through its own gatekeeper before advancing.

See docs/skills.md for the complete inventory with standards and key capabilities per skill.

How Gatekeepers Work

Supreme Team enforces quality through five adversarial gatekeepers at two levels:

Per-phase gatekeepers (gatekeeper-design, gatekeeper-build, gatekeeper-code, gatekeeper-azure) validate work within their sub-pipeline. Each specialist's output must pass its gatekeeper before the next specialist begins.

Cross-pipeline gatekeeper (gatekeeper-admiral) validates at the boundaries between sub-pipelines — ensuring the output of one pipeline is suitable input for the next.

Every gatekeeper verdict is one of:

• APPROVED — advance to the next phase
• REVISE — return with specific findings to address (max 2 cycles)
• ESCALATE — surface the blocking issue to the user

A review that finds nothing is treated as the most suspicious review of all.

See docs/gatekeepers.md for the full pattern.

Persistent Saves

Pipeline state is automatically saved to skillset-saves/ in your project workspace as the pipeline runs. This provides:

• Cross-session resume — close the conversation, come back later, pick up exactly where you left off
• Crash recovery — lease-based locking and idempotent gatekeeper submissions prevent corruption from session crashes
• Audit trail — every state transition, gatekeeper verdict, and revision cycle is logged chronologically
• Deliverable backup — every SRS, architecture doc, test report, and security audit is saved to disk as it's produced
• Graceful degradation — if saves fail, the pipeline continues with in-context artifacts and warns you about persistence gaps

See docs/persistent-saves.md for details.

Limitations

• Context window dependent — large projects can exceed an AI assistant's context window. The save system mitigates this with reference-mode tiers, but very large codebases may still require manual chunking.
• LLM accuracy — the pipeline enforces structure and adversarial review, but the quality of outputs is bounded by the underlying model's capabilities. Gatekeepers catch many issues but are not infallible.
• Azure-specific — the cloud deployment sub-pipeline currently supports Azure only. Other cloud providers would require additional skill sets.
• No runtime execution — Supreme Team generates artifacts (code, configs, runbooks) but does not execute deployments automatically. The deployer skill produces commands and scripts; a human or CI system runs them.
• Single-session concurrency — the lease-based lock system is advisory. Running two sessions against the same project simultaneously can cause conflicts.

Quick Start

See QUICK-START.md for installation steps and first-use instructions.

Documentation

Document	Description
QUICK-START.md	Installation and first-use guide
Install.md	Detailed installation procedure (AI-agent and manual)
AGENTS.md	Authoritative skill manifest for tool discovery
docs/architecture.md	Pipeline architecture, flow diagrams, and execution modes
docs/skills.md	Complete skill inventory with standards
docs/gatekeepers.md	Gatekeeper pattern and adversarial philosophy
docs/persistent-saves.md	Save system, resume, and audit trails
docs/direct-invocation.md	Standalone skill usage and fallback prompts
docs/directory-structure.md	Repository and installed layout reference

---

_{Built by TykoDev · Supreme Team}