开源 · MIT 许可 · 使用你自己的 API 密钥运行

不要把调用哪个模型写死在代码里。

两个核心概念：策略决定每次调用由哪个模型处理；工作流将这些决策串联成完整的处理管道。开源、MIT 许可，使用你自己的提供商 API 密钥运行，无需信任任何中间层。

策略 · 基本单元

选择模型的是策略，不是你。

把 model="gpt-5.5" 写死在代码里，它很快就会过时。新模型每周发布，价格持续变动，部分请求需要工具调用能力、图像理解，或者只是需要更低成本的选项。策略声明资格条件和排名方式，针对每次请求选出合适的模型，并在失败时执行回退。你传入的是规则，而不是一个名称。

每次决策都会留下可重放的追踪记录：相同的策略、输入和模型目录，可以在宿主机或你自己的机器上精确还原决策过程。没有任何路由发生在暗处。

组合路由调用

工作流对整个任务进行路由。

真实的功能不是单次调用。一个客服回复包含分类 → 起草 → 审核：每个步骤需要不同的模型，审核环节可以在内容发出前拒绝它，各步骤可以并行展开再汇合。一条策略路由一次调用；工作流路由整个处理管道，并生成一份完整的统一追踪记录。

低成本分类，达到质量基线后起草，再经过强力 no-log 审核——可在发出前拒绝内容。

查看 flow_ir

flow.support-ticket.json

["flow", {
  "u": {"kind": "input"},
  "t": {"kind": "llm", "system": "Classify the ticket and extract the account id as JSON.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["has_cap", "supports_json_mode"]],
      ["neg", ["normalize", ["field", "price_out"]]], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u"]},
  "d": {"kind": "llm", "system": "Write a reply using the ticket and the triage.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["cmp", "bench_intelligence", "ge", 0.55]],
      ["neg", ["normalize", ["field", "price_out"]]], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u", "t"], "template": "Ticket:\n$1\n\nTriage:\n$2"},
  "g": {"kind": "llm", "system": "Check brand voice, PII, refund limits. Refuse if any fail.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["is", "no_log"]],
      ["field", "bench_intelligence"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["d"]},
  "out": {"kind": "output", "inputs": ["g"]}
}]

低成本起草，强力批评，再改写。在预算内实现质量跃升。

查看 flow_ir

flow.draft-critique-revise.json

["flow", {
  "u": {"kind": "input"},
  "d": {"kind": "llm", "system": "Draft an answer.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["neg", ["normalize", ["field", "price_out"]]], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u"]},
  "c": {"kind": "llm", "system": "Critique the draft: list concrete flaws and gaps.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["field", "bench_intelligence"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["d"]},
  "r": {"kind": "llm", "system": "Rewrite the answer, fixing every point in the critique.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["field", "bench_intelligence"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u", "d", "c"], "template": "Question:\n$1\n\nDraft:\n$2\n\nCritique:\n$3"},
  "out": {"kind": "output", "inputs": ["r"]}
}]

从同一强力策略中进行 N 次随机抽样（sample 将它们分散到排名靠前的候选模型），再由裁判择优选出结果。

查看 flow_ir

flow.best-of-n.json

["flow", {
  "u": {"kind": "input"},
  "n1": {"kind": "llm", "system": "Answer the question.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["field", "bench_intelligence"], ["sample", 0.5], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u"]},
  "n2": {"kind": "llm", "system": "Answer the question.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["field", "bench_intelligence"], ["sample", 0.5], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u"]},
  "n3": {"kind": "llm", "system": "Answer the question.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["field", "bench_intelligence"], ["sample", 0.5], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["u"]},
  "j": {"kind": "llm", "system": "Pick the single best candidate; return it verbatim.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["field", "bench_intelligence"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["n1", "n2", "n3"], "template": "A:\n$1\n\nB:\n$2\n\nC:\n$3"},
  "out": {"kind": "output", "inputs": ["j"]}
}]

三个按族系固定的模型并行起草，第四个模型综合出更优的单一答案。

查看 flow_ir

flow.panel.json

["flow", {
  "u": {"kind": "input"},
  "a": {"kind": "llm", "system": "Draft an answer.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["family_eq", "gemini-3.1-pro-preview"]],
      ["zero"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]], "inputs": ["u"]},
  "b": {"kind": "llm", "system": "Draft an answer.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["family_eq", "claude-opus-4-8"]],
      ["zero"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]], "inputs": ["u"]},
  "c": {"kind": "llm", "system": "Draft an answer.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["family_eq", "deepseek-v4-flash"]],
      ["zero"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]], "inputs": ["u"]},
  "f": {"kind": "llm", "system": "Synthesize the single best answer from the drafts.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["add", ["scale", 0.7, ["normalize", ["field", "bench_intelligence"]]],
             ["scale", 0.3, ["neg", ["normalize", ["field", "price_in"]]]]],
      ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["a", "b", "c"]},
  "out": {"kind": "output", "inputs": ["f"]}
}]

推理模型和编码模型在不同过滤条件下并行运行，再经合并步骤融合输出，适用于推理与代码混合的任务。

查看 flow_ir

flow.specialist-split.json

["flow", {
  "u": {"kind": "input"},
  "rz": {"kind": "llm", "system": "Reason through the problem step by step.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["is", "cap_reasoning"]],
      ["field", "bench_intelligence"], ["argmax"], ["id"], ["always", {"action": "next_candidate"}]], "inputs": ["u"]},
  "cd": {"kind": "llm", "system": "Produce any code the problem needs.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]], ["cmp", "bench_coding_rank", "le", 5]],
      ["add", ["scale", 0.7, ["normalize", ["field", "bench_coding"]]],
             ["scale", 0.3, ["neg", ["normalize", ["field", "price_out"]]]]],
      ["argmax"], ["id"], ["always", {"action": "next_candidate"}]], "inputs": ["u"]},
  "m": {"kind": "llm", "system": "Merge the reasoning and the code into one answer.",
    "policy": ["policy", ["and", ["meets_req"], ["not", ["is", "disabled"]]],
      ["add", ["scale", 0.6, ["normalize", ["field", "bench_intelligence"]]],
             ["scale", 0.4, ["neg", ["normalize", ["field", "price_out"]]]]],
      ["argmax"], ["id"], ["always", {"action": "next_candidate"}]],
    "inputs": ["rz", "cd"], "template": "Reasoning:\n$1\n\nCode:\n$2"},
  "out": {"kind": "output", "inputs": ["m"]}
}]

横向滑动查看图示 →

在文档中查看所有工作流模式 →

开始使用

两个仓库，MIT 许可，现在即可使用。

克隆 Host 和 Engine 仓库，配置你的提供商账号，今天就可以用自己的 API 密钥开始路由。无加价，无隐藏路由逻辑。

unhardcoded

参考宿主实现：接收策略并生成指纹，使用你的提供商 API 密钥路由调用，并写入追踪记录。

查看仓库 →快速开始 →

unhardcoded-engine

policy_ir 和 flow_ir 规范、参考解释器，以及其他实现用于自我验证的一致性向量。

查看仓库 →policy_ir →

MIT 许可提供商中立可移植 IR 可自托管

不想自己运行宿主？托管版本（含托管目录、追踪存储和团队管控功能）即将上线。

或发送邮件至 hello@unhardcoded.com 加入。

不发垃圾邮件。托管版就绪时仅发一封通知邮件。