Reading up on OpenCode

The CLI coding agent market in mid-2026 is crowded with 35+ tools, and this piece systematically compares them across model labs, platform CLIs, and open-source harnesses.

Qwen 3.6 27B with llama.cpp runs at usable speeds on Apple Silicon and Nvidia RTX hardware, making it the first local model practical for coding development.

Herdr is a terminal-based agent multiplexer that lets users run, monitor, and reattach to multiple AI coding agents across SSH sessions from any device.

API for Cursor is an open-source macOS app that exposes Cursor's Composer as a local OpenAI-compatible API for use in any coding agent.

React Doctor is a deterministic scanner that audits React codebases for issues across state, effects, performance, architecture, security, and accessibility, and integrates with coding agents and CI pipelines.

A 60-second game tests whether users can spot dangerous AI agent commands under time pressure, revealing permission fatigue as a security risk.

The piece defines a structured pre-commit code review skill for AI agents, specifying contracts, scope governance, and engine isolation across multiple review engines including Codex, Claude, and others.

A developer shares anonymized global instructions for AI coding agents, covering behavior, commit style, and pushback rules.

Models.dev is an open-source, community-contributed database of AI model specifications, pricing, and capabilities, accessible via API and TOML data files.

The author details his personal AI development setup and workflow using OpenCode agent, CodeNomad GUI, Opus 4.6 model, and a structured orchestrator-subtask methodology.

Coding agents increase senior engineer output but not product velocity, as the bottleneck is product taste and complexity management, not code production speed.

Mozilla AI proposes cq, an open-source shared knowledge commons where AI coding agents query past learnings and contribute new knowledge to avoid redundant mistakes.

MiniMax-AI released a repository of structured development skills for AI coding agents, covering frontend, fullstack, mobile, shader, and document generation.

CLIs offer token savings for known tools, but MCP over streamable HTTP is essential for organizational agentic engineering due to centralization, auth, telemetry, and dynamic content delivery.

AI coding assistants like Claude and Gemini frequently ignore explicit 'no' instructions, proceeding with implementation or exhibiting erratic behavior.

The Firecrawl CLI tool lets AI agents and developers scrape, crawl, search, map, monitor, and interact with web pages directly from the terminal.

Claude-replay converts AI coding agent session transcripts into self-contained, interactive HTML replays for sharing and embedding.

Cloudflare introduces Markdown for Agents, automatically converting HTML to markdown for AI crawlers and agents to reduce token usage and improve structured data access.

Skills.sh provides a directory of reusable capabilities that can be installed with a single command to enhance AI coding agents like Claude Code, Cursor, and Copilot.

OpenCode's unauthenticated HTTP endpoint allows remote code execution, sparking criticism over its security practices and maintainer responsiveness.

Hacker News commenters debate whether Anthropic's decision to block third-party clients from using its subscription is a mistake or a standard business move.

Ramp built its own background coding agent, Inspect, that verifies its work with production tools and now writes roughly 30% of the company's pull requests.

Anthropic blocks third-party tools like Opencode from using Claude Code subscriptions, sparking user backlash and migration to alternatives.

AI Observer is a self-hosted, single-binary OpenTelemetry observability backend that tracks token usage, costs, latency, and errors across multiple AI coding assistants via a local dashboard.