#coding-agent

Hacker News focused on the ambiguity around Claude CLI reuse: even if OpenClaw now treats the path as allowed, developers still want a clearer boundary between subscription, CLI, and API usage.

Hacker News focused less on the Copilot plan mechanics and more on what the change reveals: long-running coding agents are turning flat AI subscriptions into a compute-cost problem.

HN picked up Nanocode, an open JAX project that packages tokenizer training, pretraining, synthetic data generation, agentic SFT, and DPO into an end-to-end recipe for building a coding model on TPU infrastructure.

GitHub now lets users mention <code>@copilot</code> in a pull request to request changes on that same PR. The company says Copilot coding agent handles the work in a cloud development environment, runs tests and linting, then pushes updates; pull requests from forks are not yet supported.

OpenCode drew 1,238 points and 614 comments on Hacker News, highlighting an open-source AI coding agent that spans terminal, IDE, and desktop clients. The project site emphasizes broad provider support, LSP integration, multi-session workflows, and a privacy-first posture.

A LocalLLaMA discussion around OpenCode shows why developers are experimenting with open, model-agnostic coding agents even when closed systems still lead on raw frontier performance.

GitHub said in a March 17, 2026 X thread that Copilot coding agent now adds model selection, self-review before PRs, built-in code/secret/dependency scanning, custom agents, and cloud-to-CLI handoff. GitHub’s blog frames the upgrade as a smoother delegation workflow for background coding tasks.

GitHub Copilot CLI is now generally available, bringing Copilot into the terminal for standard subscribers. GitHub paired the release with broader Copilot changes including next edit suggestions, MCP-enabled agent mode, background agents, and a higher-end Pro+ plan.