What it is and who makes it
A terminal-native coding agent with interactive and scripted surfaces, provider configuration and an open-source workflow.
Maker or project: OpenCode project. Classification: Open-source terminal coding agent; this page treats the surrounding context, tools, permissions and verification as the relevant harness boundary.
Start the terminal UI in a project.
Use the configured provider and agent mode to inspect, plan or change files.
Run non-interactive commands for bounded automation where the exit and output contract is clear.
Review diffs and tests as the acceptance boundary.
Who it may fit
Pick an agent by its control surface and trust boundary, not only by its model label.
Useful scenarios
- Open-source and terminal-native workflows
- Developers who want to switch providers or models
- Interactive sessions that can later be scripted or served
Trade-offs and checks
- OpenCode project identity, releases and commands can move quickly; verify the official docs and repository before scripting it.
- Headless or server modes need authentication and network controls, just like any internal service.
- Model/provider portability does not guarantee identical tool-call or structured-output behavior.
Interactive vs headless and CI
These modes change the approval, identity and observability boundary.
The TUI is interactive. The documented run, server/web and protocol surfaces support scripts or headless clients; configure authentication, binding, permissions and machine-readable output rather than scraping terminal text.
Request-to-verification workflow
Use this mental model when debugging a surprising edit, tool call or cost.
Intent + context
Your prompt, repository instructions, selected files, history and environment become model context.
Tool + permission
The harness proposes a read, edit, shell, web, MCP or other action; policy decides whether it can run.
Result + verification
Command output, compiler errors, tests and your review return to the loop before acceptance.
Setup outline
Start with the smallest safe experiment, then follow the current official documentation for the real installation.
# Follow the current official install guide curl -fsSL https://opencode.ai/install | bash cd your-project opencode # Non-interactive workflow opencode run "List the failing tests; do not edit files."
Instructions, skills, MCP and hooks
These surfaces can shape context or add actions; they are part of the trust boundary.
OpenCode has project/global configuration and documented agent, MCP, plugin and server surfaces. Check which instruction file and permission settings apply to the current directory before running a shared repository configuration.
Official sources
Use these sources for current installation, capabilities, plans, releases and security notices.
Reviewed 2026-09-26. This date is a change-tracking note, not a promise that the product remains unchanged.
Related learning
Related agent guides
Want the foundations behind the workflow?
Study AI fundamentals, prompt design, AI-assisted development, agents and MCP without a forced linear course.