Stats
Actions
Tags
From oh-my-agent
Checks or sets oma reasoning effort profile (low/medium/high/xhigh) for planning depth, verification strictness, and cost/delay control. Outputs status, effort matrix, and recommendations.
How this command is triggered — by the user, by Claude, or both
Slash command
/oh-my-agent:reasoningThe summary Claude sees in its command listing — used to decide when to auto-load this command
运行 oma 推理努力管理器。
输入：
$ARGUMENTS
协议：
1. 解析请求的努力值。
2. 支持的值：
- `low`
- `medium`
- `high`
- `xhigh`
3. 如果未提供值，从 `.oma/state/reasoning.json` 显示当前配置文件（如果可用）。
4. 将努力映射到 oma 姿态：
- 规划深度
- 验证严格性
- 预期延迟/令牌成本
5. 如果文件系统工具可用，写入/更新 `.oma/state/reasoning.json`。
6. 澄清边界：此配置文件控制 oma 编排行为和建议，而不是直接控制 iFLOW CLI 运行时标志。
建议的 `reasoning.json` 格式：
{
"effort": "low|medium|high|xhigh",
"planning_depth": "minimal|balanced|deep|maximum",
"verification_strictness": "light|normal|high|very-high",
"updated_at": "<ISO-8601>",
"source": "oma:reasoning"
}
输出格式：
## 推理状态
- 当前：
- 请求：
- 已应用：
- 已持...运行 oma 推理努力管理器。
输入： $ARGUMENTS
协议：
lowmediumhighxhigh.oma/state/reasoning.json 显示当前配置文件（如果可用）。.oma/state/reasoning.json。建议的 reasoning.json 格式：
{
"effort": "low|medium|high|xhigh",
"planning_depth": "minimal|balanced|deep|maximum",
"verification_strictness": "light|normal|high|very-high",
"updated_at": "",
"source": "oma:reasoning"
}
输出格式：
| 努力 | 规划深度 | 验证 | 成本/延迟 |
|---|
npx claudepluginhub sswym/oh-my-agent --plugin oh-my-agent/model-routeRecommends the optimal LLM tier (haiku, sonnet, opus) for the current task based on complexity, risk, and budget. Accepts optional task description and --budget flag.
/configConfigures Plan-Build-Run settings including model selection, workflow toggles, depth, gates, and parallelization.
/model-routeRecommends the best model tier (haiku, sonnet, opus) for a task based on complexity, risk, and budget constraints, with optional --budget flag.
/rigorSelects rigor profile (lean, standard, thorough, exhaustive, custom, inherit) for quality-vs-tokens tradeoff and persists to global or project config. Interactive if omitted.
/opusSwitches current session to Claude Opus behavioral mode for maximum reasoning depth on complex architecture decisions, deep analysis, critical code review, and similar tasks.
/planCaptures user intent, analyzes requirements, and generates a weighted execution plan saved to .claude/session-plan.md. Does not execute — plans are saved for review or later execution via /octo:embrace.