🤓 Latest Submissions
ShipSafe — AI Release Readiness System
Every software team knows the moment. It is late Friday afternoon. The code is written, the pull request is merged, and CI is green. But the Release Captain — the person who must say yes or no — is not sure. The security scan produced 847 warnings. The test report says 61% coverage. What does any of that mean? Is it safe to ship? So teams do one of two things: delay the release and lose business value, or guess and risk a production incident over the weekend. According to the State of DevOps Report, release failures and deployment anxiety are the top sources of engineering team stress. The root cause is not a lack of tools — teams already have SonarQube, Snyk, and Jest. The root cause is that those tools give long lists and leave the decision to humans who cannot interpret 847 warnings in 15 minutes. ShipSafe replaces the list with a decision. It is an AI-powered release readiness system built on IBM Bob 2.0 that gives every project a Release Readiness Score from 0 to 100 and a definitive RED or GREEN verdict. Anyone — a CTO, a PM, a new developer — understands the answer in five seconds. ShipSafe works through a team of four IBM Bob custom modes. The Security Agent scans for OWASP Top 10 vulnerabilities: SQL injection, path traversal, hardcoded secrets, broken authentication, and IDOR. The Quality Agent measures cyclomatic complexity, detects dead code, and checks error handling completeness. The Test Agent analyses not just coverage percentages but test completeness — whether the right things are tested, including security rejection tests. The Captain Mode orchestrates all three as parallel IBM Bob subagents, applies a transparent weighted formula (Security 40%, Quality 35%, Tests 25%), and delivers the final verdict. Critically, ShipSafe does not just report — it acts. After scanning, the Captain automatically applies safe fixes: replacing hardcoded secrets with environment variables, adding JWT token expiry, patching error detail leakage, updating .gitignore, and generating 15 critical security test cases. Issues requiring human judgment are listed with estimated fix times and IBM Bob guidance commands. The live demonstration takes a real Node.js API from 38 (RED — DO NOT SHIP) to 92 (GREEN — SAFE TO SHIP) in under two minutes, producing a beautiful HTML release report showing exactly what was fixed and what remains. ShipSafe directly addresses the release and deployment workflow identified in the hackathon challenge. Its target users are Release Captains, CTOs, and senior engineers at any company that ships software. It is unique because no existing tool — not SonarQube, not Snyk, not any CI gate — combines multi-agent analysis with automatic remediation and a single human-readable verdict.
27 Sep 2026
