Gitar launches AI-code validation platform
A startup developer infrastructure company has built AI agents for code review and CI workflows. The company is called Gitar, and its platform addresses the challenge of keeping up with AI-generated code validation using only manual qualilty gates, which cannot scale.
As the amount and pace of code generated by AI continues to increase, so do the number and size of pull requests and code changes. The result is teams have to pass code without proper quality checks and engineers find themselves working on pipeline failures.
“The industry has focused on accelerating code generation, but the real constraint is shipping that code safely,” said Ali-Reza Adl-Tabatabai, co-founder and CEO of Gitar. “Developers today spend too much time acting as the integration layer between CI failures, logs, fixes, and approvals. Gitar turns that process into an autonomous system that reviews and quality checks code, triages problems, diagnoses root causes, and proposes fixes so engineers can focus on delivering software.”
Gitar’s platform enables AI code review , CI failure root cause analysis by de-duping and summarizing the failure to find the issue, and provide agents that help developers query, update and fix code directly inside pull or merge requests, according to the company announcement. Further, the platform allows teams to create custom agents for automated workflow validation, and connects to such tools as GitHub, GitLab, Jira, Slack, CircleCI and more.
Workflow analytics help teams ensure CI reliabliity and code quality by tracking validation outcomes and identifying recurring issues, which creates opportunities to improve development workflows.
Cloudflare Agent Week
Here is a recap of announcements from Cloudflare’s Agents Week. The announcements incldue expanding the developer toolset. You can explore the full breakdown on Cloudflare Radar.
-
Launched Agent Cloud, a new compute model for the agentic era: Agent Cloud gives agents a secure, scalable home that persists across long-running tasks, with infrastructure including Dynamic Workers (100x faster and a fraction of the cost of traditional containers) and expanded model access including OpenAI’s GPT-5.4 and Codex.
-
Announced general availability of Sandboxes, first introduced last June: A Sandbox is a persistent, isolated environment powered by Cloudflare Containers. Since launch, Cloudflare added secure credential injection, PTY support, persistent code interpreters, and increased limits alongside active CPU pricing, improving support for running agents at scale.
-
Expanded Enterprise MCP with new governance tools designed to make agentic workflows safer and simpler: Updates include Code Mode with MCP server portals to reduce token costs associated with MCP usage, and integration with Cloudflare Gateway for Shadow MCP detection to identify unauthorized remote MCP servers.
-
Introduced Project Think, the next generation of the Agents SDK: While current agents are constrained by short lifespans, many intended tasks are long-running and cross-platform. Project Think is designed to support persistent, long-running agents within the Agents SDK.
-
Added real-time voice capabilities to the Agents SDK through an experimental voice pipeline: The framework is designed for scale, enabling developers to integrate voice into existing agent architectures so it can evolve alongside other agent capabilities.
- New tool to track the agentic web: Cloudflare launched isitagentready.com, a tool that audits any URL to see if it’s actually prepared for AI agents (like which standards they’ve adopted from robots.txt signals, content permissions, and integration endpoints). Data on the top 200,000 domains suggests most of the web is predominantly “read only” for agents. You can explore the full breakdown on Cloudflare Radar.
The post SD Times News in Brief appeared first on SD Times.