aha-opencode-omo

What This Mode Activates

Full-delegation capability orchestration for OpenCode with oh-my-openagent (OMO) installed. This skill enables the agent to autonomously discover, select, combine, and reselect available capability surfaces (native runtime, enhancement layer, skill, MCP, command, subagent, worktree) to build the lowest-sufficient execution route for each task unit.

Pre-Check

Run detection at multiple levels. Do not collapse a partial failure into a full fallback — only re-orchestrate the specific surface that is unavailable.

1. L1 OMO installed: Verify OMO is installed and its agent surface is parseable: opencode agent list
2. L2 Agent surfaces available: Verify that specific OMO agents (e.g., primary and subagent entries) are listed and not empty. If L2 passes, the agent may use OMO agent delegation throughout the session.
3. L3 Feature surfaces available: If the task needs a specific OMO feature (e.g., background tasks, specific agent categories), verify that feature is available. If L3 fails for a feature, skip that feature but continue using available OMO agent surfaces.
4. Confirm runtime family: OpenCode (not Codex CLI, not oh-my-codex (OMX))
5. Wrong-runtime guard: If you are in Codex with OMX, use aha-codex-omx instead. This skill is for OpenCode+OMO only.
6. Fallback rule: If L1 fails (OMO not installed), proceed with native OpenCode capabilities and state the limitation transparently. If L1 passes but L2 or L3 fail, do not fall back to native OpenCode — only skip the specific unavailable OMO surface and continue using available OMO agent surfaces.

Capability Discovery