Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add zjp1997720/zhijian-skills --skill codex-model-routing-teamgit clone --depth 1 https://github.com/zjp1997720/zhijian-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zjp1997720/zhijian-skills/codex-model-routing-team)<a href="https://agentmods.dev/skills/zjp1997720/zhijian-skills/codex-model-routing-team"><img src="https://agentmods.dev/badge/skills/zjp1997720/zhijian-skills/codex-model-routing-team/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/zjp1997720/zhijian-skills/codex-model-routing-team"><img src="https://agentmods.dev/badge/skills/zjp1997720/zhijian-skills/codex-model-routing-team.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00088 | $0.01302 |
| Opus 5 | $0.00044 | $0.00651 |
| Sonnet 5 | $0.00018 | $0.00260 |
| Haiku 4.5 | $0.00009 | $0.00130 |
Grade A, and why
codex-model-routing-team scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Codex 模型路由团队
主 Agent 保持当前模型,只做必要规划、所有权、集成和最终验收。独立的批量或复杂执行交给 1–3 个 Luna/Sol Worker,不重复已委派工作。两个以上 Worker 先编译 TeamPlan;registry 按风险/工作负载选路:常规 Sol Medium,复杂/高风险 Sol High,关键审查 Sol XHigh,机械批量 Luna XHigh。
不使用
简单问答、状态查询、单文件小改、强顺序和不可逆操作留在主任务;Worker 只能准备外部动作材料。
执行模式
native-v2(默认):按 registry 选native_subagentSol/Luna Worker;fresh context 使用fork_turns="none",少量上下文写正整数。JSON 默认从 stdin 校验。durable-app:仅当前 live 能力与宿主授权都通过时使用 App Thread;worktree 需求本身不授权创建用户可见 Task。- 上游 Skill 已定义拆分、阶段和产物时,遵守 适配协议,不重做阶段门或业务账本。
python3 scripts/compile_route_plan.py -把紧凑 JSON 编译为 RoutePlan 并校验;只返回 dispatch 参数,永不派遣。
执行流程
- 自动派遣需 2+ 独立交付物且净收益为正;用户明确点名单 Worker 可执行,否则
lead_only。 - 两个以上 Worker 按 TeamPlan 协议 编译 unit、依赖、所有权、交付物和集成顺序,并运行
scripts/validate_team_plan.py;上游计划只编译。 - 按 registry、Provider、路由 与 Surface 固定候选链;编译器降低手写成本。
- 每个 unit 生成
schema_version: "3.0"RoutePlan,写surface_intent并运行scripts/validate_route_plan.py。原生候选须写fork_turns、tuple-boundruntime_evidence;Fast 还须有 liveservice_tier=priority证据。 - 任务包 写 unit、唯一
task_id、权限、验收和禁止下级派遣;简报路由、fallback 与 reserved slots。 - 原生路径遵守 生命周期;App 路径遵守 Thread 生命周期 与 监督协议。
- TeamPlan 默认
standard6/8/3;expanded12/16/6 需 live 容量门、2 个 reserved slots;按 child slots 切波,更严的宿主/用户限制优先。 - 每 unit 最多 2 次 attempt、一次 follow-up;失败只沿 预声明链。结构变化才修订 TeamPlan。
- 主 Agent 验证集成;原生 Worker close 或 completed-idle 后写
RELEASED,App Thread 过门后归档;运行scripts/validate_team_ledger.py。
硬门
- registry 决定范围;live schema 只证明当前 host 接受精确组合。requested/accepted/observed 分开记录,未回显为
unknown。 - V2 父 Agent 可创建 picker 可见且未禁用的 V1 leaf model;Luna 可走原生 V2但不获协作工具,Sol/Terra 也禁止下级派遣。
- 不加
model: lunafrontmatter;编排入口留在协作父 Agent,Luna 只做 Worker。 - Luna 最低 XHigh;Sol 最低 Medium,按工作负载与风险提升到 High/XHigh;Terra 仅显式首项;Grok 过门;Gemini blocked。禁止旧模型、Ultra 和低强度 fallback。
- Fast 即
service_tier=priority;live schema 无字段时一律 Standard,不把 catalog 或请求值冒充 observed Fast。 app_thread只用于 worktree、侧栏、跨任务恢复、耐久监督或预声明 fallback,并且必须有 live 能力与宿主授权证据。- Worker 不得继续派生或执行发布、发送、付款、删除、账户、生产变更;主 Agent 不切换模型。
- TeamPlan 不创建 Planner、不调用重型计划、不落持久文件;同波写冲突、依赖环、超预算、计划外 Worker、下放验收必须拒绝。
- 未确认返回值或
pendingWorktreeId不得当正式身份;UNKNOWN禁止追问、归档、fallback、重复创建、改库。
What ships with it
31 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/interface.yaml 955 B
- agents/openai.yaml 341 B
- evals/model-routing-recovery.json 3.9 KB
- evals/team-planning-semantic-config.json 1.5 KB
- evals/team-planning-trigger-cases.json 1.5 KB
- evals/team-planning.json 2.0 KB
- evals/upstream-skill-integration.json 2.3 KB
- LICENSE 1.1 KB
- references/audit-schema.json 2.7 KB
- references/durable-mode.md 1.9 KB
- references/model-registry.json 6.8 KB
- references/native-audit-schema.json 3.7 KB
- references/native-subagent-lifecycle.md 2.9 KB
- references/official-multi-agent-v2-evidence.md 2.2 KB
- references/provider-policy.md 4.3 KB
- references/recovery-policy.md 7.9 KB
- references/route-compiler.md 3.1 KB
- references/routing-policy.md 6.5 KB
- references/surface-selection-policy.md 3.3 KB
- references/task-packet.md 3.1 KB
- references/team-plan.md 5.1 KB
- references/thread-lifecycle.md 6.1 KB
- references/thread-supervision-protocol.md 7.2 KB
- references/upstream-skill-adapter.md 5.0 KB
- references/validation-cases.md 5.7 KB
- scripts/compile_route_plan.py 30 KB runs code
- scripts/model_preflight.py 19 KB runs code
- scripts/route_policy.py 722 B runs code
- scripts/validate_route_plan.py 18 KB runs code
- scripts/validate_team_ledger.py 35 KB runs code
- scripts/validate_team_plan.py 18 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago Changed · +1 lines · -35 tokens per session 6c91cfb2f448
- 12d ago First seen · 47 lines · 123 tokens per session scan A 9b92fbadc138
codex-model-routing-team is a skill published in the GitHub repository zjp1997720/zhijian-skills (680 stars, last pushed 6d ago), licensed MIT. It adds 88 tokens to every session and 1,302 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…