GitHub utilized secret scanning to manage over 20,000 alerts across 15,000 repositories and achieve inbox zero.
The process involved distinguishing significant alerts from false positives and developing remediation workflows. This effort was led by Staff Security Engineer Michael Recachinas, who focuses on large-scale security programs and secure development lifecycle tooling. The initiative successfully reached inbox zero within nine months.
Kimi K2.7 Code, an open-weight model, is now generally available in GitHub Copilot as a selectable option.
This is the first open-weight model offered in the Copilot model picker, providing more choice and a lower-cost option for coding workflows. Kimi K2.
Claude Fable 5 has been re-deployed globally, featuring enhanced cybersecurity safeguards and a new AI jailbreak severity framework.
The cybersecurity safeguards include safety classifiers designed to detect and block harmful cybersecurity uses. The framework for AI jailbreak severity, developed with Glasswing partners, categorizes jailbreaks based on the extent of behaviors they unblock, from minor to highly dangerous.
This model can autonomously make plans, utilize tools like browsers and terminals, and operate at a level that previously required larger models. Claude Sonnet 5 nearly matches the performance of Opus 4.
Google released Copybara, a tool for transforming and moving code between repositories.
Copybara is used internally at Google to synchronize code between confidential and public repositories by designating one as the authoritative source. It supports contributions from any repository, and changes from non-authoritative repositories are transformed and moved to the authoritative one.
GitHub has outlined six security settings that maintainers should enable to strengthen their projects against attacks.
These settings are designed to close easy vulnerabilities, making projects more challenging to compromise. While not guaranteeing complete security, they represent a significant improvement over previous configurations. The guidance comes from Joseph, a prominent cybersecurity expert with extensive experience in educating developers on secure coding practices. His open-source initiatives and educational content have reached thousands globally, emphasizing the importance of proactive security measures.
The Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5.
Anthropic AI announced they will start restoring access to these products the next day. The company expressed gratitude to users for their patience during the period of restricted access. Further updates on the situation are expected to be shared soon.
The Godot Foundation has decided to stop accepting AI-authored code contributions for its open-source game engine.
Godot's maintainers have struggled with an overwhelming number of AI-generated pull requests, which they find draining and demoralizing. The new guidelines will prohibit AI-authored code, AI-submitted pull requests, and AI-generated text in communications. Despite the challenges, the Foundation acknowledges that the increase in pull requests reflects growing interest in Godot, but the burden on maintainers has necessitated these changes to ensure effective project management.
The OpenCode integration with Supabase is now available, allowing users to connect their agents to databases, Edge Functions, and logs.
Users can access Supabase account and project management APIs by typing /supabase and authenticating, with MCP setup configured automatically. OpenCode requires version 1.
HackerRank has open-sourced its Applicant Tracking System (ATS).
When tested, the ATS produced varying scores for the same resume, with results ranging from 74 to 90 out of 100. This inconsistency occurred despite running the same command multiple times, with only minor changes like removing debug print statements. The ATS has gained significant attention on platforms like LinkedIn and Reddit, where discussions have highlighted its unpredictable scoring behavior.