The kit is organized as an OpenCode config directory:
AGENTS.md: global behavior rules.opencode.json: models, permissions, plugin registration, default agent.tui.json: TUI plugin registration.agents/: role prompts.commands/: slash commands.docs/ai/harness/: agent-readable contracts for agents, commands, evidence, and checks.docs/ai/evolution/: AHE benchmark and evolution records.plugins/: bundled OpenCode/TUI plugins.references/: reusable checklists that local skills may cite.skills/: local skills, includingopen-design.scripts/check-harness.mjs: local harness validator.tools/: custom TypeScript tools.
The default permissions are conservative: reads are allowed, edits and bash ask by default, and external directories are denied.
The default agent is lead, acting as a bounded router for free-form messages. For small, clear, low-risk work, lead delegates direct mode to developer; for technical uncertainty it delegates to researcher; for visual or interaction work it delegates to designer; and for planning gaps it delegates to specifier. Slash commands still route to their explicit agents, for example /feature and /plan to lead, /scope to scoper, and /design to designer.
Run node scripts/check-harness.mjs from an installed OpenCode config directory to validate the core harness contracts. In this repository, npm run check runs the same validator against opencode/.
The bundled TUI token plugin is registered in tui.json. It shows lead token
usage and total usage across child/subagent sessions when OpenCode exposes that
session tree through the TUI plugin API.