From DevOps to DevSecOps
What DevOps and CI/CD automate, how DevSecOps weaves security tasks into every stage of the lifecycle, why 'shifting left' is cheaper, and the main families of automated security testing: SAST, DAST, SCA, and secret scanning.
The Problem with Manual Releases
The movie app now has a frontend, a backend, and a database. Every change carries risk: a typo can crash the server, an updated package can introduce a vulnerability, a rushed fix can expose data. When testing and checking happen by hand — if they happen — they’re inconsistent, slow, and skipped exactly when people are under pressure.
DevOps is the practice of breaking down the wall between the people who build software (development) and the people who run it (operations), and automating the path from a code change to running software. The automation backbone is CI/CD:
- Continuous Integration (CI) — every change pushed to the shared repository is automatically built and tested, so problems are caught within minutes of being introduced.
- Continuous Delivery — every change that passes is automatically prepared for release; a human approves the final step.
- Continuous Deployment — every change that passes goes to production automatically, with no manual gate.
DevSecOps: Security in Every Stage
Traditional security was a gate at the end: build the thing, then hand it to a security team for review. That fails in a world of daily releases — the gate either becomes a bottleneck or gets skipped.
DevSecOps integrates security into each stage of the DevOps lifecycle and automates it wherever possible, making security a shared responsibility rather than one team’s final checkpoint.
| Stage | Security tasks | In the movie app |
|---|---|---|
| Plan | Threat modeling, impact analysis | ”What happens if the database credentials leak?” |
| Develop | Finding and classifying vulnerabilities, security tools in the editor, automated fixes | Linting for injection risks while writing routes |
| Code commit | Dependency management, secure CI/CD pipelines, secret scanning | Flagging a vulnerable package version or a committed password |
| Build, test, deploy | Configuration validation, infrastructure scanning, application scanning | Scanning the running API for missing security headers |
| Operate and monitor | Log analysis, anomaly detection | Alerting on a spike of failed requests |
Why “Shift Left”?
Picture the lifecycle as a timeline running left to right. Shifting left means moving security checks earlier. The reason is economics: a flaw caught in the editor costs a minute to fix; caught in code review, an hour; caught in production after a breach, potentially millions — plus users’ trust. Early feedback is also better feedback, because the developer still remembers what they were doing.
Shifting left doesn’t mean only left. Some issues only appear in a running system, so monitoring and runtime scanning stay on the right. Good programs check at every stage.
The Families of Automated Security Testing
| Type | What it examines | Finds | Blind spots |
|---|---|---|---|
| SAST — static application security testing | Source code, without running it | Injection patterns, unsafe functions, hard-coded secrets | Runtime configuration, deployment issues; can be noisy |
| DAST — dynamic application security testing | The running application, from the outside | Missing security headers, exposed endpoints, injectable inputs | Code it can’t reach; needs a running target |
| SCA — software composition analysis | Third-party dependencies | Packages with known vulnerabilities (CVEs), license problems | Flaws in your own code |
| Secret scanning | Commits and history | API keys, passwords, tokens | Secrets that are encoded or split |
| IaC scanning | Infrastructure configuration | Open ports, public storage, overly broad permissions | Application logic |
A modern Node.js app often has hundreds of transitive dependencies, so SCA tends to find the most issues by volume. The next lesson uses a well-known DAST tool, OWASP ZAP, inside the CI pipeline.
OWASP — the Open Worldwide Application Security Project — is a nonprofit community that publishes free security tools and guidance, including the widely used OWASP Top 10 list of the most critical web application risks.
Common mistake: concluding that because the scanner passed, the app has no vulnerabilities.
Each tool only sees what it’s designed to see. A DAST scan won’t find a vulnerable dependency that’s never exercised; SAST won’t notice a database left open to the internet; none of them understand your business logic (“can a user see another user’s data?”). Automated checks raise the floor — they don’t replace threat modeling, code review, and testing the things that matter most to your users.
1. What distinguishes continuous deployment from continuous delivery?
Delivery keeps every passing change ready to release behind a human approval; deployment removes that gate.
2. Which testing type would most likely flag that the API responds without important security headers?
Headers are a property of the running application's responses, which DAST observes from the outside.
3. A package your API depends on has a newly published CVE. Which practice catches this?
Software composition analysis checks dependency versions against databases of known vulnerabilities.
4. Why is 'shifting left' cost-effective?
The later a flaw is found, the more work has been built on top of it — and in production, the cost can include a breach.
Up Next
Theory into pipeline: the next lesson reads a real GitHub Actions workflow — triggers, jobs, runners, steps, and secrets — and walks through what an OWASP ZAP scan does and how to read its results.