Apiiro reports finding 2,539 credentials matching its secret-classification rules in 1,994 public GitHub repositories, including values saved after Claude Code users approved commands with 'Yes, and don't ask again.' The research explains how settings, history and agent transcripts can retain credentials and recommends rotation, repository exclusions and command-aware scanning.
Why I recommend it: Apiiro did not use or test the credentials, so 'real secret' does not mean confirmed valid. Anthropic's Head of Claude Code said the team would prototype mitigations, including basic secret scanning.
DeepSeek's open-source research prototype of a proactive coding agent for a single GitHub repository, packaged for DeepSeek Harness. DeepSeek labels it reference research code, not an official or production-ready product.
Jovi De Croock describes an open-source agent-native Git platform organized around human-approved plans, evidence-based review lanes, merge queues and reversible releases. The article discusses prototype limitations and proposed development practices rather than independently verified production results.
Free, open-source Git hook that pauses protected-branch pushes to ask questions about what the changed code does. Supports plain Git and several coding-agent workflows. The software is free; use of an external model endpoint may incur separate costs.
#code review#coding agents#open source
sphinx-six-theta.vercel.appAdded Oct 19, 20260 opens
Explains that Google Antigravity now lists Claude Opus 5.5 and Sonnet 5.5 for non-trial Google AI Pro and Ultra subscribers, and that Claude 4.6 and GPT-OSS-120b are removed on November 2, with who is affected and what to do.
An open-source agent skill on GitHub by nicobailon that generates rich HTML pages or slide decks for diagrams, diff reviews, plan audits, data tables, and project recaps.
Simon Willison's annotated slides and notes from his closing keynote at WeAreDevelopers World Congress North America (Sept. 2026): a chronological tour of the year's LLM developments, starting with the November 2025 models that made coding agents start working.
A hands-on write-up of wiring Google's open Gemma 4 model into the Codex command-line coding agent so it runs locally instead of calling a hosted API.
From the site: I wanted to know whether Gemma 4 could replace a cloud model for my day-to-day agentic coding. Not in theory, in practice. I use Codex CLI…
Why I recommend it: Useful if you want to try coding agents without paying per token — local models are slower, but free and private.
A free, open-source AI coding agent for VS Code, JetBrains, the command line and the cloud, with support for local models and your own API keys at no markup.
From the site: Kilo is the open source AI coding agent for VS Code, JetBrains, CLI, and Cloud. Access 500+ models, bring your own keys at zero markup, and keep code private with local models.
Why I recommend it: Open source and free to install — you only pay a model provider if you choose a hosted one.
