Google Cloud expands CodeMender and AI agent security tools
Google Cloud has added sandboxed validation to its CodeMender vulnerability-fixing agent and made Auth Manager generally available to secure AI agents.
SUMO· the number puzzle
Dialo· learn languages by speaking
ediviewer· read EDI files like documents
giftdetect· the right gift in minutesGoogle Cloud on November 5 (US time) announced a set of security upgrades aimed at the era of AI agents. They include new features for CodeMender, an agent that finds and fixes software vulnerabilities, along with stronger agent authentication and threat detection and response.
The company argues that agentic coding tools have sped up software development so much that human review cannot keep pace with machine-speed commits and unverified dependencies, which could open backdoors for attackers. Its answer is real-time guardrails in the integrated development environment (IDE) and model checks that remove flaws before code is merged.
CodeMender gets sandboxed validation
CodeMender uses several AI models, including Gemini 4 Argon, to identify unknown vulnerabilities and to generate and verify patches. It has been in preview since July. The new version can run vulnerability checks and attack simulations in isolated environments, using OS-level sandboxes or virtual machines, with human confirmation built in for file writes and command execution.
Users can now limit scans to a specific folder, an entire repository, or only files changed since the last scan, which reduces verification time and cost.
Agent credentials and Wiz integration
Auth Manager, which centrally manages credentials for AI agents, is now generally available. It stores credentials in Google-managed storage and automates OAuth 2.0/PKCE flows and token refresh, lowering the risk of credential theft. Agent Identity gains preview support for JWT: by issuing standard OIDC tokens, it lets agents authenticate to AWS, Azure, on-premises systems and MCP servers without long-lived static API keys.
Google Cloud is also deepening its ties with Wiz to protect autonomous applications from development through runtime. Wiz AI SAST, in preview, uses AI to analyze application behavior and detect complex business-logic flaws, and it can hand fixes to tools such as CodeMender. Wiz Defend and Google Security Operations now share a common data model, allow investigations from either product and sync findings in both directions. Wiz Defend is also integrated with Google Threat Intelligence for direct access to the latest threat data.
Apps worth a look
SUMO: the number puzzleTap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.
Dialo: learn languages by speakingTalk to an AI conversation partner; every sentence comes back corrected and explained in your language. 9 languages, A1 to C2.
ediviewer: read EDI files like documentsOpen EDIFACT and X12 messages, see every segment explained, convert to JSON/XML, export PDF and Excel. Runs locally, no upload.
giftdetect: the right gift in minutesPick who it is for, the occasion and their personality. The search engine reads product descriptions from local shops and ranks what fits. Free, no sign-up.
More from this day
See all →
Citus Demonstrates GPS-Free Firefighter Tracking, Eyes Military Use
Operation Ghostbusters: 830,840 Suspected Ghost Students Found
JNSA Security Quiz Service Hit by Breach, 15,128 Records at Risk
