OpenCode

ship

Autonomous coding agent that edits files, runs tests, and validates changes in your repo.

Visit opencode.ai →Add to kit Open in Noemium

ledger entry · observed by @whysanesanders

Spins up a full agent loop in your project, writes code, runs tests, and commits when it succeeds. Good for well-scoped tasks with existing test coverage. Still rough on large refactors and can hallucinate library APIs.

Field guide: OpenCode →

Strengths

Full agent loop in your repo: edit, test, commit when it succeeds.
Open-source path for well-scoped tasks that already have tests.

Known limitations

Agent loops can wander on ambiguous requirements.
Token consumption on large codebases adds up fast.
Less mature editor integration than Cursor or Windsurf.

Use it when

Bounded coding jobs with existing test coverage.
A terminal agent when you do not want a full AI IDE.

Skip it when

The task is an ambiguous large refactor.
Token burn on a big repo would dominate the bill.
You need mature editor integration like Cursor.

Buy-check

·Price — freemium, free tier; paid plans from $20/mo as of 2026-08-19. Open the pricing page, not a screenshot.
·Free tier — Free tier exists — burn it first.
·Cancel fallout — What happens to your data/projects on cancel — check the ToS, not the FAQ.
·Seat math — Per-seat pricing multiplies quietly. Count seats before the annual plan.
·Who skips it — Also skipped when: The task is an ambiguous large refactor.; Token burn on a big repo would dominate the bill..