- Home
- Skills
- Agents & Orchestration
- sdlc multi agent harness
sdlc multi agent harness
Deterministic SDLC Multi-Agent Harness (5-Role Pipeline + JSON Contracts)
$29
Works with the AI Agents you already use
sdlc multi agent harness
Example session with this skill installed
Build a secure Node.js Express microservice for user registration. Include JWT auth, a Postgres schema, unit tests, and a Dockerfile. Apply the full SDLC harness.
- Read your context and instructions
- Compiled the sdlc multi agent
I have orchestrated the development of your User Registration Microservice. The Architect designed a layered service pattern, the Developer implemented the Express logic with bcrypt hashing, the QA Reviewer verified the JWT implementation, and DevOps provided the containerization.
User Registration Microservice
Architectural Overview
- Pattern: Layered Architecture (Controller -> Service -> Repository)
- Stack: Node.js, Express, PostgreSQL, Prisma ORM, Vitest
- Security: Argon2 password hashing, JWT-based stateless authentication
Directory Tree
user-service/
├── src/
│ ├── controllers/
│ ├── services/
│ ├── middleware/
│ └── index.ts
├── tests/
├── prisma/
│ └── schema.prisma
├── Dockerfile
├── docker-compose.yml
└── README.md
Source Code: Authentication Middleware
import { Request, Response, NextFunction } from 'express';
import jwt from 'jsonwebtoken';
export const authMiddleware = (req: Request, res: Response, next: NextFunction) => {
const token = req.headers.authorization?.split(' ')[1];
if (!token) return res.status(401).json({ error: 'Unauthorized' });
try {
const decoded = jwt.verify(token, process.env.JWT_SECRET as string);
(req as any).user = decoded;
next();
} catch (err) {
res.status(401).json({ error: 'Invalid token' });
}
};
Security Audit Summary
- Status: PASSED
Findings: Verified Argon2 implementation for password entropy. Confirmed JWT secret is pulled from environment variables.
- Tests: 12 unit tests passed (100% coverage on auth logic).
DevOps & Deployment
FROM node:20-slim
WORKDIR /app
COPY package*.json ./
RUN npm install
COPY . .
RUN npx prisma generate
EXPOSE 3000
CMD ["npm", "start"]
Next steps
- Add your
DATABASE_URLandJWT_SECRETto the.envfile. - Run
docker-compose up --buildto launch the service and database. - Execute
npm run testto verify the security suite in your environment.
Connects securely to your tools. The creator never sees your data.
About this skill
Deterministic SDLC Multi-Agent Harness
Transform your AI coding assistant into a deterministic, zero-trust software delivery pipeline. Deterministic SDLC Multi-Agent Harness coordinates 5 isolated specialist roles using formal JSON Schema contracts, Blind Security Auditing, and an automated Self-Healing Fix Loop.
Built for Claude Code (sub-agents), Cursor, Windsurf, Codex CLI, Gemini CLI, Cline, and GitHub Copilot.
Why It Outperforms Standard Prompts
- 5 Isolated Specialist Roles (
agents/*.md):- Orchestrator: Extracts requirements, enforces context hygiene, and assembles the final deliverable without writing code directly.
- Architect: Produces a concrete tech stack, directory tree, data models, interface contracts, and security strategy—no "TBD" placeholders.
- Developer: Implements every file in the blueprint verbatim with explicit error handling and zero
TODO/passstubs. - Blind QA & Security Reviewer: Audits code against OWASP Top 10 / CWE and generates runnable tests without seeing Developer notes.
- DevOps & Docs Engineer: Generates non-root multi-stage
Dockerfile,docker-compose.yml,.env.example, and an accurateREADME.md.
- True Blind Security Review: Strips all Architect and Developer reasoning before QA/AppSec auditing to eliminate confirmation bias.
- 5 Formal JSON Schema Contracts (
schemas/*.schema.json): Validates every internal delegation envelope and specialist output (HandoffEnvelope,ArchitectOutput,DeveloperOutput,QaSecurityOutput,DevOpsDocsOutput). - Self-Healing Fix Loop: Automatically routes
criticalandhighsecurity or contract defects back to the Developer for surgical patches (up to 2 iterations) before final delivery. - Minimum Context Rule: Sub-agents receive strictly the structured inputs required for their phase—never bloated chat histories.
What’s Included in the Package
SKILL.md— Master Orchestrator pipeline, handoff rules, and context isolation protocol.agents/architect.md— Principal Software & System Architect role specification.agents/developer.md— Senior Full-Stack Software Engineer role specification.agents/qa-security-reviewer.md— Principal QA Automation & AppSec Reviewer specification.agents/devops-docs.md— Senior DevOps & Technical Writer specification.schemas/*.schema.json— 5 JSON Schema (Draft 2020-12) contracts enforcing structured agent outputs.examples/worked-microservice-example.md— End-to-end execution walkthrough showing JSON handoffs, a Blind Security Review catch (SEC-001,BUG-001), and a Self-Healing Fix Loop resolution.
Quick Start
- Extract into your agent's skills directory (e.g.,
~/.claude/skills/sdlc-multi-agent-harness/or.claude/skills/sdlc-multi-agent-harness/), or load on-demand via Agensi MCP. - Invoke with your project specification:
- "Run sdlc-multi-agent-harness to architect, implement, security-audit, and containerize a rate-limited FastAPI microservice."
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Capped at 2 self-healing fix loops. Requires environment support for sub-agent tool calls or isolated context steps. Sub-agents cannot access external live internet resources directly.
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 this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
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.
Trust & safety
Security scan
Security scan passed today