AI Coding Assistants

Today's Stories

1. 1

   IBM Db2 Genius Hub 1.1.5 Adds CLI and Gemini Models

   IBM launched Db2 Genius Hub 1.1.5 with a new Db2 Genius CLI that lets developers and DBAs ask database questions in plain English from the terminal, plus log-based alerting, HashiCorp Vault integration for credentials, predictive graphs in AI conversations, and Gemini models via Google Cloud Vertex AI. These features bring AI assistance directly into existing developer and DBA workflows, which may help teams identify issues earlier, manage credentials centrally, and choose preferred enterprise AI models without switching tools.

   The shift from reactive troubleshooting to proactive database management hinges on whether teams adopt the new CLI and alerting schedule, and whether organizations using Vertex AI actually select Gemini alongside Claude.

2. 2

   NTTドコモ builds robot ML pipeline in 3 days with AI-DLC

   NTT Docomo's MIT Division joined AWS's AI-DLC Train The Trainer and built a robot-arm pipeline — imitation learning (IL) refined by reinforcement learning (RL) — in 3 days, using the RPD (Refined Policy Distillation) approach. The Inception phase alone surfaced 11 inconsistencies and produced 30+ structured documents, suggesting AI-DLC can cut rework in ML projects where design is uncertain and knowledge is scattered.

   Whether the framework holds in parallel team development is untested — the article's second part reports on Team Pink and Team Blue running the pipeline side by side.

3. 3

   Unity ships Claude Code and Codex plugins to fix stale AI skills

   Unity released official plugins for Claude Code and OpenAI's Codex, giving agents skills its own teams write and maintain. The Codex version launches with 31 skills covering interfaces, 2D graphics, URP, audio, and in-app purchases. According to Unity, general-purpose agents typically pick up forum posts and tutorials for outdated engine versions whose code may compile but often doesn't work as intended.

   The plugins run on Unity 6 and up, so older projects may not be covered. In Codex they install with a click; in Claude Code via npm.

4. 4

   AI bug hunting drives 66,401 CVEs as slowdown talks stall

   A wave of AI-assisted vulnerability discovery has hit the software industry. Microsoft issued patches for 974 CVEs in one month, a new record, while cve.icu has logged 66,401 CVEs so far this year. AI is surfacing more known flaws, not necessarily more total flaws, which is putting pressure on under-resourced IT and security teams and open source volunteers to keep pace with patching.

   The balance hinges on whether defenders can use AI as effectively as attackers, says Cisco's Matthew Olney. Watch whether patch adoption keeps up with discovery.

5. 5

   Oracle's Clay Magouyrk: AI coding rollout outpaced our processes

   Co-CEO Clay Magouyrk said Oracle last year hadn't found broad internal use for generative AI; after rolling out ChatGPT Enterprise and Codex, adoption hit 80% in three months, with GPT-6 Astra costing 2.5 times more than other models. Coding that once took a team two to three quarters now takes about a week, but Oracle says that speed hasn't shortened delivery to customers because testing and release processes haven't kept pace.

   The test is whether Oracle redesigns testing, validation, deployment, and release management fast enough to convert faster coding into faster customer delivery; watch the roughly 60% to 70% false-positive rate on Anthropic's Mythos Preview.

6. 6

   HarnessRouter standardizes agent runs via one protocol

   HarnessRouter, an open-source implementation of the Unified Harness Protocol, runs Codex and Claude Code through one Responses-style API, currently at version 2026-09-12. Teams can pick or swap the agent runtime per feature without rebuilding progress, session, file, and cancellation handling in their product.

   UHP standardizes the task lifecycle but does not make harnesses behave the same, so identical prompts can still produce different plans and tools. Local sessions use separate OS users, not separate containers.