explore | Grilled Cheese

El gateway normaliza, aplica policy, pide aprobación humana, audita y recién ahí ejecuta. Falla el audit log, se cancela la llamada. Fail closed, no fail open. #ciberseguridad #devsecops #iagentica

Stop manually correlating CVEs and KEV data. Feed 290,000+ threat IPs to your firewall, enrich SIEM alerts, and block vulnerable CI/CD builds with working API code. #DevSecOps #SIEM #Security

https://www.valtersit.com/guides/api/valtersit-cve-api-threat-intelligence-integration-guide/

A severe zero-click vulnerability named Plugin4Shell is threatening AI coding assistants, bypassing key security verification checks.

https://pcmasterinsider.com/plugin4shell-ai-coding-agent-vulnerability-github-copilot/

#Cybersecurity #AICoding #GitHubCopilot #DevSecOps

Stop guessing what "GPL-ish" means. Grant groups licenses by risk so you can approve/deny in seconds. One list, not fifty rules.

👉 https://anchore.com/blog/grants-release-0-3-0-smarter-policies-faster-scans-and-simpler-compliance/

#OpenSource #SupplyChainSecurity #Compliance #DevSecOps

Last trap: a Kyverno PolicyException must also list the autogen- rule name, or the Deployment stays blocked.

Full workflow, policy, and break-glass:
https://devtocash.com/blog/2026-09-19-cosign-keyless-signing-github-actions-sbom-kyverno-verify
#DevSecOps #Kubernetes

A new AI supply chain vulnerability called #Plugin4Shell highlights a serious weakness in how AI coding agents handle plugins.

#Cybersecurity #AISecurity #ArtificialIntelligence #SupplyChainSecurity #DevSecOps #GitHub #ClaudeCode #OpenAI #Copilot #SoftwareSecurity

thehackernews.com/2026/09/plug...

OWASP's 2026 LLM Top 10 reveals where AI security risks really live—and it's not always in the model. Incident data shows the bigger threat is in the systems around it.

https://www.resilientcyber.io/p/the-2026-owasp-llm-top-10-and-the

#appsec #DevSecOps