reasoning

Control omi reasoning effort level to balance depth, rigor, latency, and token cost.

Purpose

Use this skill to set an explicit reasoning posture before planning, execution, or verification-heavy turns.

Trigger

  • User asks for faster vs deeper reasoning
  • Session needs tighter quality gates or lower latency
  • Team wants predictable planning/verification effort across stages

Supported Effort Levels

  • low
  • medium
  • high
  • xhigh

Workflow

  1. Identify requested effort or infer medium as default.
  2. Explain impact on planning depth, verification strictness, and cost/latency.
  3. Persist/update .iflow/state/reasoning.json when filesystem tools are available.
  4. Recommend a next command (/omi:mode, /omi:team, or /omi:loop) based on posture.

Output Template

## Reasoning
- requested:
- applied:

## Profile
| Dimension | Value |
| --- | --- |

## Next Command
- ...