Review · AISource: Automater News
OpenCode 2.0 unifies clients around one local process
OpenCode 2.0 connects terminal, web, desktop, and script clients to one local process, Automater reports. The process starts automatically. [1] The official homepage identifies Anomaly in its copyright notice; DiffVibe traces the team to Serverless Stack. [2][3] Automater reports that 2.0 adds asynchronous subagents and shell commands. It also reports Code Mode for scripting Model Context Protocol (MCP) tools. [1] DiffVibe reports a rebuilt server API and says existing 1.x plugins are incompatible with 2.0. [2]
AI coding agent
OpenCode 2.0
VerdictPromising
Measured improvements in coding accuracy, task cost, or reliability remain unestablished. Automater reports support for work across multiple interfaces. [1]
- Best for
- Reported workflow: shared terminal, web, desktop, and script clients, according to Automater [1]
- Pricing
- Agent software: free under MIT; model usage can incur separate charges. Optional Zen access is prepaid and usage-based; its page lists a $20 balance purchase plus a $1.23 processing fee. [5][6][7]
Pros
- Shared local process across clients, according to Automater [1]
- Asynchronous subagents and shell commands, according to Automater [1]
- Free MIT-licensed agent software, according to Csorba [5]
- Embeddable SDK and plugin hot reloading, according to Automater [1]
Cons
- Incompatibility with existing 1.x plugins, reported by DiffVibe [2]
- SDK still changing, according to Automater’s account of the announcement [1]
- Separate charges possible for model usage [5]
Why it matters
Automater’s account points to a practical benefit: a common local process for multiple interfaces. Subagents and shell commands can run in the background. [1] In his v1 review, Balázs Csorba describes broad provider support. He recommends Claude Code for Claude-centered workflows and Aider for smaller, Git-centered workflows. [5] Migration deserves attention: Aiola’s October 13, 2026, changelog documents fixes for the new API and server authentication. [4]
Attribution
Sources
- 1Automater NewsOpenCode 2.0 Runs One Background Service for Every Client and Ships Code Mode, Background Subagents, and an Embeddable SDK
- 2DiffVibeOpenCode 2.0 ships a shared service, Code Mode, and a rebuilt API
- 3OpenCodeOpenCode | The open source AI coding agent
- 4AiolaAiola Changelog — Grok CLI, OpenCode & GUI Lifetime
- 5Balázs CsorbaOpenCode review: open-source agent, any model
- 6OpenCodeOpenCode Zen | A curated set of reliable optimized models for coding agents
- 7OpenCodeZen | OpenCode
Read the original source for full details and context.
AI
Anthropic moves Claude Opus 5.5 fast mode out of research preview
Anthropic removed access requests and the beta-header requirement for Claude Opus 5.5 fast mode. Default fast mode rate limits cover Start, Build and Scale, and Azure-hosted Microsoft Foundry deployments are supported.
AI
AWS adds four managed Airflow agent skills; upgrade skill supports Provisioned only
AWS added four agent skills for coding assistants to author, debug and test managed Airflow workflows and upgrade environments. Authoring, debugging and testing support Provisioned and Serverless; the upgrade skill supports Provisioned only.
Review · AI
Elastic’s Linux profiling preview: Setup, requirements and limits
Elastic says its OpenTelemetry Profiling integration collects continuous CPU profiles on Linux through eBPF. It requires no application instrumentation or restarts, according to Elastic. Version 0.0.1 is a technical preview. OpenTelemetry Profiles remains Alpha. Elastic’s October 14, 2026, walkthrough performs the one-time Universal Profiling storage initialization as superuser. It then configures a Fleet policy with samples_per_second=20 and enrolls the Linux agent. To verify collection, Elastic directs users to Infrastructure > Universal Profiling > Flamegraphs. Users select the traffic window and filter process.executable.name: "java"; agent health alone is insufficient. Elastic reports 30,150 PetClinic requests in four minutes at slightly over half a CPU core. The walkthrough does not report a profiling-overhead measurement.