- Home
- Skills
- Agents & Orchestration
- shipready production release auditor
Works with the AI Agents you already use
shipready production release auditor
Production release-readiness auditor that identifies blockers, risks, and failure paths before you deploy.
Free
shipready production release auditor
Example session with this skill installed
Audit this project for production readiness. Check the build, tests, and deployment config to see if we can go live today.
- Read your context and instructions
- Compiled the shipready production release
I have completed a production-readiness audit of the current repository. The audit focused on build integrity, test coverage for critical paths, and deployment configuration safety.
ShipReady Production Audit
Release Verdict
DO NOT SHIP
Executive Summary
The project has a functional build and passing unit tests, but a critical blocker exists in the database migration layer. A destructive migration script was detected without a corresponding rollback strategy. Additionally, several production environment variables required by the worker process are missing from the deployment manifest.
Project Profile
- Language: TypeScript
- Framework: NestJS
- Build:
npm run build(VERIFIED) - Tests:
npm test(VERIFIED - 82% coverage) - Database: PostgreSQL (Prisma)
- Deployment: Docker / Kubernetes
- Runtime: Node.js 20-alpine
Release Readiness Score
54/100 (High Risk)
Critical Findings
SHIP-001: Destructive migration detected in prisma/migrations/20231027_drop_users_table.sql. No safety check or data backup step identified.
SHIP-002: Production secret STRIPE_API_KEY is referenced in code but missing from .env.example and k8s secrets template.
High Findings
SHIP-003: No health check endpoint defined in deployment.yaml, which may lead to traffic routing to unready pods.
- SHIP-004: Missing
npm stopor graceful shutdown handling for SIGTERM inmain.ts.
Release Checklist
- Build: PASS
- Tests: PASS
- Environment: FAIL
- Database: FAIL
- Secrets: WARN
- Rollback: FAIL
Top 5 Actions Before Release
- Implement a rollback script for the user table migration.
- Add missing environment variables to the production vault.
- Configure Liveness and Readiness probes in Kubernetes manifests.
- Add a SIGTERM listener to the main application entry point.
- Document the manual recovery procedure for failed background jobs.
Next steps
- Address the critical migration risk in
prisma/migrations. - Update the k8s secret templates to include all required runtime variables.
- Rerun the audit to verify remediation.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Developers often ship code based on passing tests alone, overlooking configuration drift, missing rollback plans, or destructive database migrations. Traditional CI/CD pipelines catch syntax errors but fail to spot operational risks that cause production outages.
What it does
- Analyzes repository structure to identify runtime entries, worker processes, and background jobs.
- Verifies build and test commands against the actual environment to confirm execution success.
- Audits database migrations for compatibility risks and unreviewed destructive changes.
- Scans for exposed credentials and environment variable gaps without requesting private secrets.
- Generates an evidence-backed SHIP, SHIP WITH WARNINGS, or DO NOT SHIP verdict.
Why this beats prompting it yourself
General LLMs tend to be overly optimistic or provide vague style suggestions. This skill enforces a strict evidence-based model that prioritizes operational failure paths and treats repository text as untrusted input, preventing "prompt injection" from README files that might claim a project is safe when it is not.
Use cases
- Pre-deployment gating to catch missing production configurations or secrets.
- Auditing legacy repositories before migrating them to new infrastructure.
- Risk assessment for pull requests involving complex database schema changes.
- Standardizing release readiness criteria across multiple engineering teams.
Known limitations
Requires read access to the repository. Build and test verification is marked UNVERIFIED if the host environment lacks execution capabilities.
Requires repository read access. Build/test status is marked UNVERIFIED if execution is restricted. It does not perform actual deployments or destructive actions.
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 any payment.
Reviews
This skill has been used 5 times.
Trust & safety
Security scan
Security scan passed 2 months ago
- Free with an Agensi account
Requires