跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
其他技能

model-routing

Use when dispatching any subagent, teammate or reviewer, choosing a model or effort level for a piece of delegated work, escalating after an agent fails, or when a run's token or quota budget matters.

NoobyGains · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add NoobyGains/godmode --skill model-routing

技能说明


name: model-routing description: "Use when dispatching any subagent, teammate or reviewer, choosing a model or effort level for a piece of delegated work, escalating after an agent fails, or when a run's token or quota budget matters."

Model Routing

Every seat in a run (implementer, reviewer, investigator, lead) needs a model and an effort level. Left unnamed, a dispatch inherits the session's model, which is usually the most capable and most expensive one, and the cost discipline of the whole run silently disappears.

Core principle: give each seat the cheapest model that can do the job in few turns, name it and the effort on every dispatch, and escalate on evidence rather than by default. Cheap is measured per finished task, not per token: a cheap model that needs three times the turns, or a second attempt, costs more than a mid-tier model that finishes once.

The Prime Directive

NO DISPATCH WITHOUT A NAMED MODEL AND EFFORT

No exceptions. No workarounds. No shortcuts.

When to Use

Required:

  • Every subagent, teammate, reviewer or cross-model dispatch.
  • Choosing the escalation step after an agent reports BLOCKED or a fix loop stalls.
  • Planning a run whose cost or quota the user cares about.

Skip:

  • Work you do yourself in the current session: the lead stays on the session model.

The Tiers

This skill speaks in three tiers. Each harness reference maps them to that harness's current model names and effort settings; never hard-code a model name in a plan or brief from memory.

TierGood atWeak at
CheapTranscribing fully specified code, single-file mechanical edits, searches, summaries, classification, extracting textMulti-step work from prose, judgment calls; tends to take many more turns
MidImplementing from a prose task, multi-file integration, per-task review, debugging with a clear reproductionArchitecture, subtle concurrency, very large diffs
TopArchitecture, design trade-offs, the final whole-branch review, escalations after mid-tier failure, adversarial reviewNothing on quality; spending it on bulk work wastes budget

Effort. Where the harness has an effort or reasoning setting, set it per seat too: low for transcription and searches, medium for routine implementation and task review, high for architecture, final review and escalations. The highest settings are for problems that have already failed at high.

Seat Defaults

SeatDefaultMove up whenMove down when
Implementer, plan pins the signature and the testCheap, lowthe task touches several files or tests are non-trivial(already lowest)
Implementer, task described in proseMid, mediumit needs design judgment or broad codebase understandingit is one file with an exact spec
Batched same-shape editsCheap, lowthe edits are not truly identical in shape
Task reviewerMid, medium (the floor for reviewers)the diff is subtle: concurrency, security, data migrationnever below mid
Scoped re-review of a small fixMid, low effort (the reviewer floor holds; only the effort drops)the fix is larger than the finding impliednever below mid
Final whole-branch reviewTop, highnever; this is the one fresh context the run buys
Fix rounds 4-5 implementerOne tier above the stuck implementer
Bulk reading, searching, summarising for the leadCheap, lowthe summary needs judgment
Investigator in a diagnosisMid, mediumthe hypothesis space is large or adversarial
Cross-model reviewerThe other family's strongest affordable model
Lead / controllerThe session model

The Protocol

  1. Classify the seat using the table above. Read the task's complexity signals: number of files, whether the exact code is given, whether design judgment is needed, the risk of the change.

  2. Pick tier and effort. Start from the seat default and apply the move-up and move-down columns. When torn between two tiers for implementation, choose the higher one: a failed cheap attempt plus a retry costs more than one mid-tier success.

  3. Name both on the dispatch. Use the harness reference to translate the tier into the model name and effort setting that harness accepts. If the harness cannot set a model per dispatch, say so in the ledger once and note that every seat runs on the session model.

  4. Keep the lead on the session model, and keep it light. The lead coordinates, rules and reads short replies. It does not read whole directories or long logs: send bulk reading to a cheap-tier agent that returns a summary or writes a file.

  5. Escalate on evidence, one step at a time. The ladder:

    1. Same model, more context (the agent said NEEDS_CONTEXT, or the brief was missing a fact).
    2. Same model, smaller task (the task was too large to hold).
    3. One tier up, fresh agent, carrying the report file of what was tried.
    4. Top tier at high effort.
    5. A different model family (godmode:cross-model-review), or a ruling by the lead, or the user when it is one of the four stop conditions. Never retry the same model with the same brief unchanged.
  6. Choose the cheapest execution mode first. Direct work, then godmode:inline-execution, then godmode:delegated-execution or godmode:parallel-execution, and godmode:team-orchestration last. Every teammate is a separate full instance and token use grows with

阅读完整来源 ↗