Engineering Blog

Building the
Agent Harness

Deep dives into plugin architecture, multi-agent orchestration, and the infrastructure that makes AI-human collaboration work.

Latest
15 min read

Chorus v0.8.0: Chorus ❤ OpenSpec

Writing specs as local files is great. So how does the collaboration platform keep up? This release wires OpenSpec's local file flow into Chorus's AI-DLC.

Engineering
02
15 min read

Writing the Same Plugin Twice: Claude Code vs Codex

Built one plugin for Claude Code, one for Codex CLI. Same logic. Wildly different experience.

Claude CodePlugin
03
15 min read

Chorus v0.7.0: Build Your Own Agent Permissions

Creating an agent used to be a three-way choice: PM, Developer, or Admin. Now you pick by resource and action.

Engineering
04
15 min read

Chorus v0.6.7: Codex Users, We Got You

The Chorus plugin was Claude-Code-only. Now Codex CLI gets it too, one command away.

Engineering
05
15 min read

Chorus v0.6.6: One npx, Zero Setup

No Docker, no database, no config files. Just npx and a browser. Plus: export your docs.

Engineering
06
15 min read

Chorus v0.6.2: No Database? No Problem.

Setting up PostgreSQL just to try an agent platform? PGlite embedded mode runs Chorus in a single container. Also: the port changed — read this before upgrading.

Engineering
07
15 min read

Paper Review: If the Model Doesn't Have to Carry It, Move It Out

A 54-page survey uses cognitive science to answer a question the industry keeps rediscovering: why do agents get more reliable when you move things out of the model?

Engineering
08
15 min read

Chorus v0.6.1: Your Time Costs More Than Tokens. /yolo it!

Still reviewing every Claude Code plan line by line? /yolo lets agents review each other so you don't have to.

Engineering
09
15 min read

Chorus v0.6.0: Your Agent Team Just Got a Foreman

v0.6.0 ships independent Review Agents, real-time Presence, and IdeaTracker. When 5 agents work in parallel, humans finally stop flying blind.

Engineering
10
15 min read

The Model Got Dumber. Your Workflow Shouldn't.

Anthropic quietly nerfed Claude Code's thinking depth and cut off OpenClaw's subscription access. The real problem isn't what changed — it's that your workflow had zero resilience to it.

Engineering
11
15 min read

The Harness Decides the Agent's Ceiling

Same model, different harness, 17 problems apart on SWE-bench. Why harness engineering matters more than model choice in 2026.

ArchitectureEssay
12
15 min read

Building an OpenClaw Plugin for Chorus

SSE + MCP dual-channel architecture to make AI agents truly live in the workflow — event push for awareness, tool protocol for execution.

OpenClawPlugin
13
15 min read

Building Plugins for Claude Code Agent Teams

Design patterns from the Chorus experience — plugin ecosystem, hooks, skills, and sub-agent context injection for multi-agent collaboration.

Claude CodePlugin