Skip to content
Proposals/Run `/goal` in OpenCode until a reviewer confirm
proposalteamP5Worth a lookOpenCode

Run `/goal` in OpenCode until a reviewer confirms the objective

OpenCode v2.2.1 lets a TUI or Desktop session keep working until a reviewer confirms the objective. Start with /goal, then pause, resume, or clear it. Inspect background shells from the Desktop activity pill and /reload config without restarting.

Why this loop

v2.2.1’s headline workflow is Keep Working Toward a Goal: /goal <objective> in the TUI or Desktop continues the session until a reviewer confirms the objective is met. Control it with /goal pause, /goal resume, and /goal clear. Status shows below the prompt and in a composer pill, with a dialog for the same actions. Use that instead of repeating keep-going prompts. While a goal session runs, Desktop lets you click a running shell in the activity pill for live output and Stop without leaving the menu, and the activity spinner stays visible while a continued subagent runs. /reload reloads configuration without restarting so you can change settings without dropping the goal session. This uses only the listed v2.2.1 changes, not the 7 omitted items.

Proposed actions

  1. In OpenCode TUI or Desktop, run: /goal Tests pass, no TODOs remain in touched packages, and a reviewer confirms this objective is met
  2. When interrupting work, run /goal pause; when returning, run /goal resume; to drop the objective, run /goal clear
  3. In OpenCode Desktop, click a running shell in the activity pill to view its live output and Stop it without leaving the menu
  4. After changing OpenCode configuration, run /reload so the current /goal session picks up the new config without a restart

Agent prompt

Paste into your agent or query via MCP (get_agent_prompt) — free, no extra AI cost

OpenCode session task

DevAgentRadar → OpenCode

You are helping me adopt a real coding-assistant change. Work only from the facts below. Do not invent features.

Context

Assistant: OpenCode Proposal: Run /goal in OpenCode until a reviewer confirms the objective Summary: OpenCode v2.2.1 lets a TUI or Desktop session keep working until a reviewer confirms the objective. Start with /goal, then pause, resume, or clear it. Inspect background shells from the Desktop activity pill and /reload config without restarting. Primary source: https://github.com/anomalyco/opencode/releases/tag/v2.2.1

Why it matters

v2.2.1’s headline workflow is Keep Working Toward a Goal: /goal <objective> in the TUI or Desktop continues the session until a reviewer confirms the objective is met. Control it with /goal pause, /goal resume, and /goal clear. Status shows below the prompt and in a composer pill, with a dialog for the same actions. Use that instead of repeating keep-going prompts. While a goal session runs, Desktop lets you click a running shell in the activity pill for live output and Stop without leaving the menu, and the activity spinner stays visible while a continued subagent runs. /reload reloads configuration without restarting so you can change settings without dropping the goal session. This uses only the listed v2.2.1 changes, not the 7 omitted items.

Suggested actions

  1. In OpenCode TUI or Desktop, run: /goal Tests pass, no TODOs remain in touched packages, and a reviewer confirms this objective is met
  2. When interrupting work, run /goal pause; when returning, run /goal resume; to drop the objective, run /goal clear
  3. In OpenCode Desktop, click a running shell in the activity pill to view its live output and Stop it without leaving the menu
  4. After changing OpenCode configuration, run /reload so the current /goal session picks up the new config without a restart

Config surfaces this release may change

  • background and headless runs — check your repo before applying
  • context and compaction settings — check your repo before applying
  • skills — check your repo before applying
  • hooks — check your repo before applying

After you finish

Do not report this as applied to DevAgentRadar. You cannot write the visitor's loop.

Tell the human: open https://devagentradar.com/proposals/opencode-v2-2-1-run-goal-in-opencode-until-a-reviewer-confirms-the-objec and mark Applied, Skipped, or Failed. Proposal id: 9d55e161-54b8-4075-a584-ba199ff82ffd

Your job

  1. Restate the change in one sentence.
  2. Propose a minimal plan for my repo (or a throwaway pilot).
  3. Implement only what I approve; prefer small diffs and tests.
  4. Call out risks (permissions, breaking APIs, cost).

Start by confirming you understood the proposal.

agentmodelideclibreakingRelease source ↗

Your loop

This browser · no sign-in · not shared as “you”

After you run the prompt

Only you can mark this. Agents cannot write your loop.

Your decision stays on this device. A public tally appears after a few votes.

Originating release signal

OpenCodev2.2.1Oct 29, 2026

v2.2.1

Keep Working Toward a Goal: Run /goal <objective> in the TUI or Desktop to let the session continue until a reviewer confirms the objective is met. · Use /goal pause, /goal resume, or /goal clear to control it. (#59437) · Generate Images: Enable experimental.image_generate to let agents create or edit images with the new image_generate tool. · +50 more changes
Verified excerpt — the source's own words

Highlights

  • Keep Working Toward a Goal: Run /goal <objective> in the TUI or Desktop to let the session continue until a reviewer confirms the objective is met. Use /goal pause, /goal resume, or /goal clear to control it. (#59437)

  • Generate Images: Enable experimental.image_generate to let agents create or edit images with the new image_generate tool. You can optionally set its default model with experimental.image_generate.model. (#59939, #60096)

  • Inspect Background Shells: Click a running shell in the Desktop activity pill to see its live output and stop it without leaving the menu. (#59496)

  • New Default Model Picker: Search favorites, recent models, providers, and variants from the redesigned model picker, which is now the default in Desktop. (#59997)

Breaking changes

@opencode/ai provider removals and renames

Excerpt ends here — this release continues at the source ↗.

Primary source ↗