Your own code
Injection, broken access control, insecure cryptography and business-logic flaws in the code your team writes — found while the change is still in review.
VamiAppSec runs static code analysis with Semgrep and an LLM review for business-logic flaws, scans direct and transitive dependencies with Grype and Trivy and catches secrets and IaC misconfigurations — on every commit, as SARIF in the IDE and the pull request, with quality gates that block critical findings before they are merged.
Most vulnerabilities enter a product long before it runs. Four sources need to be checked on every change:
Injection, broken access control, insecure cryptography and business-logic flaws in the code your team writes — found while the change is still in review.
Direct and transitive libraries with known vulnerabilities, matched against NVD, OSV, EPSS and CISA KEV.
API keys, tokens and certificates committed by accident — caught in pre-commit and CI before they ship.
Misconfigurations in infrastructure-as-code templates and container images, checked with Checkov and Trivy.
Each scan type runs at the pipeline stage where it fits — and every result lands in the same deduplicated backlog.
| Scan type | How VamiAppSec runs it | What you get |
|---|---|---|
| SAST code stage | Semgrep rules on every commit, plus a semantic review by the LLM for business-logic flaws. | Findings as SARIF in the IDE and the pull request |
| Secrets pre-commit and CI | Gitleaks in pre-commit hooks and in the pipeline. | Hits with age, exposure and rotation guidance |
| SCA build stage | Grype and Trivy match direct and transitive dependencies against NVD, OSV, EPSS and KEV. | Dependency findings with VEX statements |
| IaC and containers build stage | Checkov and Trivy check infrastructure code and container images. | Misconfigurations found before deployment |
| Enrichment LLM | Each finding gets a plain-language explanation, an exploitability assessment and a fix written for your stack. | Fix suggestion right in the pull request |
| Tracking and evidence DefectDojo | Deduplicated and fingerprinted in DefectDojo, synced to Jira, Azure DevOps Boards or ServiceNow and mapped to OWASP ASVS 5.0. | ASVS coverage per chapter and level |
Gates sit where a change is still cheap to fix. Critical findings stop the merge; regressions soft-fail, so teams keep shipping.
Pre-commit hooks run secrets scanning; SAST results show up as SARIF directly in the IDE.
The PR gate runs SAST and secrets scanning; findings and a drafted fix appear as comments in the pull request.
The build gate runs SCA, IaC and container scans and checks the SBOM for completeness.
Everything lands deduplicated in DefectDojo or your tracker — with SLAs, risk acceptance and an audit trail.
Block on critical findings, soft-fail on regressions — defined as policy-as-code.
German data centers operated by Deutsche Telekom — EU jurisdiction, GDPR by design.
Scanners, DefectDojo and the LLM on your own cluster, air-gapped if needed — your code never leaves your network.
The LLM review can run on Qwen3.8 in Sankt Augustin or on your own GPUs — no third-country API, and your code is never used for AI training.
Every LLM call is logged with model version, prompt and output, and fix suggestions go through your normal code review — the model advises, your developers decide.
SAST analyses the code your team writes for vulnerabilities such as injection or broken access control. SCA checks the open-source and third-party components you use for known vulnerabilities. VamiAppSec runs both in the same pipeline and tracks the results in one backlog.
Semgrep for SAST with a semantic review by the LLM, Grype and Trivy for SCA, Gitleaks for secrets and Checkov and Trivy for IaC and containers. You don't have to operate them yourself — VamiAppSec orchestrates them.
Rules find known patterns. The LLM review looks for business-logic flaws that rules cannot express and enriches every finding with a plain-language explanation, an exploitability assessment and a fix written for your stack.
If you want them to: the PR gate can block merges on critical findings while regressions only soft-fail. Results are attached to the pull request as SARIF.
Native integrations for GitLab, GitHub and Azure DevOps — pipeline templates, PR/MR checks and quality gates — plus Bitbucket and Jenkins. Results are also available as SARIF and via a REST API.
SAST, DAST and review findings are mapped to OWASP ASVS 5.0 requirements, with coverage per chapter and verification level L1–L3 — useful for customer assessments and audits.
We connect a sample repository, run SAST, SCA, secrets and IaC scans and show you the findings in the pull request, the DefectDojo board and an OWASP ASVS coverage report.