Model Brain
Use this skill when the goal is to choose the right model for a single message or action.
Core policy
- - Prefer
zero-llm when a deterministic skill or script can do the job. - Use the cheapest adequate model for low-risk work.
- Use
bankr/claude-sonnet-4.5 for routine wallet and treasury operations that are important but still normal. - Escalate to
bankr/claude-opus-4.6 only for explicitly high-stakes, security-sensitive, or ambiguous wallet flows. - Use
bankr/gpt-5.2-codex for coding-heavy implementation work. - Keep routing deterministic and rule-based before adding any LLM-on-LLM behavior.
- Treat the output as a routing recommendation, not as transaction authority.
Default routes
- deterministic skills like
pet-me-master
- shellable tasks with no reasoning need
- casual chat, lightweight rewriting, simple summaries, low-risk classification
- general reasoning, planning, product thinking, routine wallet ops
- coding-heavy patching, repo surgery, implementation details
- long-context synthesis and broad document digestion
- lightweight vision and quick multimodal triage
- explicitly high-stakes wallet actions, security reviews, tricky architecture, final escalation
Routing workflow
- 1. Check whether the request can be handled by a deterministic skill or script.
- Classify the task:
chat, code, wallet, vision, long-context, or deterministic. - Mark whether the request is high-stakes.
- Route to the cheapest safe model.
- Return the primary route, fallback route, and a short reason.
Quick start
CODEBLOCK0
Wrapper entrypoint
Use scripts/select_model.sh when you want a simple aaigotchi-friendly wrapper before model selection. It can emit just the route, full JSON, or shell-style env lines.
References
- - Read
references/routing-table.md for the routing rules and escalation thresholds. - Read
references/bankr-models.md for the current Bankr model inventory and aaigotchi default.
模型大脑
当目标是为单条消息或操作选择正确模型时，使用此技能。
核心策略
- - 当确定性技能或脚本可以完成任务时，优先使用 zero-llm。
- 对低风险工作使用最便宜的合适模型。
- 对重要但仍属常规的钱包和资金库操作，使用 bankr/claude-sonnet-4.5。
- 仅对明确的高风险、安全敏感或模糊的钱包流程，升级至 bankr/claude-opus-4.6。
- 对编码密集型实现工作，使用 bankr/gpt-5.2-codex。
- 在添加任何LLM对LLM行为之前，保持路由的确定性和基于规则。
- 将输出视为路由建议，而非交易授权。
默认路由
- 确定性技能，如 pet-me-master
- 无需推理的可脚本化任务
- 随意聊天、轻量级改写、简单摘要、低风险分类
- - bankr/claude-sonnet-4.5
- 通用推理、规划、产品思维、常规钱包操作
- 编码密集型补丁、仓库手术、实现细节
- 长上下文综合与广泛文档消化
- 轻量级视觉与快速多模态分类
- 明确的高风险钱包操作、安全审查、棘手架构、最终升级
路由工作流
- 1. 检查请求是否可由确定性技能或脚本处理。
- 对任务进行分类：chat、code、wallet、vision、long-context 或 deterministic。
- 标记请求是否为高风险。
- 路由至最便宜的安全模型。
- 返回主路由、备用路由及简短理由。
快速开始
bash
python3 {baseDir}/scripts/route_message.py --text pet all my 53 gotchis --json
python3 {baseDir}/scripts/route_message.py --text rewrite this x thread shorter --json
python3 {baseDir}/scripts/route_message.py --text swap ETH to USDC and send to treasury --json
bash {baseDir}/scripts/select_model.sh --text build this feature in the repo --mode summary
包装入口点
当需要简单的aaigotchi友好包装器时，在模型选择前使用 scripts/select_model.sh。它可以仅输出路由、完整JSON或shell风格的环境变量行。
参考
- - 阅读 references/routing-table.md 了解路由规则和升级阈值。
- 阅读 references/bankr-models.md 了解当前Bankr模型清单和aaigotchi默认设置。
