Deterministic Dev
# Complete AI Software Engineering Suite Stop choosing between **token efficiency** for everyday coding and **strict architectural governance** for critical services. This bundle includes two battle-tested, agent-agnostic `SKILL.md` packages designed for **Claude Code, Cursor, Windsurf, Codex CLI, Gemini CLI, Cline, and GitHub Copilot**. --- ## What’s Included in This Bundle ### 1. Elite AI Team — 11-Role Tiered Engineering System (`elite-ai-team`) A fast, single-pass engineering harness engineered for developers on Pro/Plus plans who want full SDLC discipline without burning through context limits. - **Smart S/M/L Task Tiering:** Automatically activates only the roles required for a quick script (`S`), stateful API (`M`), or distributed service (`L`). - **11 Specialized Engineering Roles:** Product Owner, Architect (ADRs), Data Engineer, Security Engineer (STRIDE), Engineer, Reviewer, QA, Optimizer, SRE, DevOps, and Documentation. - **Mandatory Requirements Intake:** Batches up to 5 blocking questions first or logs explicit, versioned `ASSUMPTIONS` so nothing is silently invented. - **Resumable `STATE` Tracking + Python Validator:** Includes `scripts/validate_state.py` (zero external dependencies) to verify `STATE` blocks and catch forbidden placeholders, plus `templates/resume-session.md` for zero-drift multi-turn handoffs. - **Bundled Assets:** `SKILL.md`, `references/adr-and-stride-cheatsheet.md`, `scripts/validate_state.py`, `templates/project-context.md`, `templates/resume-session.md`, and `examples/tier-m-api-example.md`. ### 2. Deterministic SDLC Multi-Agent Harness (`sdlc-multi-agent-harness`) A modular 5-agent orchestration pipeline built for isolated sub-agent execution, strict JSON Schema contracts, and zero-trust security auditing. - **5 Isolated Specialist Roles (`agents/*.md`):** `Orchestrator`, `Architect`, `Developer`, `Blind QA & Security Reviewer`, and `DevOps & Docs Engineer`. - **True Blind Security Review:** Strips developer notes and chain-of-thought before QA/OWASP auditing to eliminate confirmation bias. - **5 Formal JSON Schemas (`schemas/*.schema.json`):** Enforces exact interface contracts, runnable test suites, non-root Dockerfiles, and zero `TODO` placeholders. - **Self-Healing Fix Loop:** Automatically routes `critical` and `high` defects back to the Developer for surgical patches (up to 2 iterations) before final assembly. - **Bundled Assets:** `SKILL.md`, 4 isolated agent specs, 5 JSON Schemas, and an end-to-end worked microservice walkthrough (`examples/worked-microservice-example.md`). --- ## When to Use Which Skill | Scenario | Recommended Skill | Why | | :--- | :--- | :--- | | Daily feature development, bug fixes, refactoring, or token-conscious builds | **`elite-ai-team`** | Single-pass execution with `S/M/L` tiering keeps context lean while enforcing ADRs, STRIDE, and tests. | | New microservices from scratch, security-critical APIs, or sub-agent pipelines | **`sdlc-multi-agent-harness`** | Enforces strict JSON handoffs, isolated context windows, Blind AppSec audit, and automated fix loops. | --- ## Quick Installation 1. Extract each skill folder into your agent's skills directory (e.g., `~/.claude/skills/` or `.claude/skills/` for Claude Code), or load on-demand via **Agensi MCP**. 2. Trigger naturally when starting a task: - *"Use elite-ai-team at Tier M to build..."* - *"Run sdlc-multi-agent-harness to architect and implement..."*
Works with any agent that signs in to Agensi, such as Claude, ChatGPT, Claude Code, Cursor and Codex CLI.
One payment, lifetime access. 2 skills unlock instantly in your library.
What's included
2 skillsSee it in action
elite ai team
Example session with this skill installed
I need a Tier M Python FastAPI service for a user profile system with PostgreSQL. Start the intake.
- Read your context and instructions
- Compiled the elite ai team
I will begin the intake for this Tier M project. Because this involves persistence (PostgreSQL) and a multi-module API, Tier M is appropriate to trigger the Architect, Data Engineer, and Security roles. ## Intake Questions 1. What is the expected peak load (requests/second) for the profile service? 2. Are there specific PII requirements or encryption standards for the user data? 3. Should the PostgreSQL schema include support for social logins or just local auth? 4. Do you have a preferred migration tool (e.g., Alembic) or should I provide raw SQL? 5. Is this service running in a specific cloud provider's serverless environment or standard Docker? ## ASSUMPTIONS (v1) - The system will use standard Bearer Token / JWT for authentication. - All profile updates must be idempotent. ```markdown # templates/project-context.md Stack: Python 3.11+, FastAPI, SQLAlchemy, PostgreSQL 15 Runtime target: Docker Environment: Dev/Staging Expected load: [PENDING INTAKE] Existing system: Greenfield Hard constraints: No external auth providers unless specified Out of scope: Frontend UI, Email sending service ``` **Next steps** 1. Answer the 5 intake questions above. 2. I will then move to the Product Owner phase to define acceptance criteria. 3. Architect phase will follow to define the component boundaries. STATE **Tier:** M **Phase:** Intake **Delivered:** Project context template, Initial assumptions **Open:** Load targets, PII requirements, Migration tool choice **Next:** Resolve intake questions
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads these skills and uses them on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every skill in this bundle goes through our trust checks.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.