OpenCode is an MIT-licensed coding agent that pioneered the skill-driven execution model. Instead of treating every request as a fresh prompt, OpenCode packages workflows into reusable SKILL.md files.
The Skill Execution Model
Every task maps to a skill. Skills live in a /skills directory with SKILL.md files containing frontmatter and step-by-step instructions. When a request matches a skill description, OpenCode loads that skill and follows its workflow.
Slash Commands
OpenCode exposes skills through slash commands: /plan, /spec, /review, /test, /ship, /debug. Each loads a specific skill with its own steps, tools, and exit criteria.
Agent Lifecycle
OpenCode maps requests to skills, loads skill instructions, executes each step, verifies outputs against exit criteria, and reports results. This structured lifecycle produces consistent output quality.
Comparison to Hermes
OpenCode is simpler and more focused. It does not have MOA, cron jobs, persistent memory, or a console dashboard. But its skill execution model is more explicit and easier to customize. For teams that want transparent, hackable workflows, OpenCode is compelling.