phocinae-server 0.1.0__tar.gz

phocinae_server-0.1.0/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 phocinae-guard authors (Nous Research)
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

phocinae_server-0.1.0/PKG-INFO

@@ -0,0 +1,196 @@
+Metadata-Version: 2.4
+Name: phocinae-server
+Version: 0.1.0
+Summary: Self-contained typed-decisions (system-one) inference server for the Phocinae-Largha-150M-v1 model
+License: Apache-2.0
+Keywords: decision-model,typed-decisions,structured-decisions,small-model,cpu-inference
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: fastapi>=0.100
+Requires-Dist: uvicorn>=0.20
+Requires-Dist: torch>=2.1
+Requires-Dist: requests>=2.25
+Provides-Extra: dev
+Dynamic: license-file
+
+# phocinae-server
+
+斑海豹（Phocinae-Largha-150M）本机决策服务：把 1.5 亿参数的决策模型打包成
+一行 `pip install` 级别依赖的 FastAPI 服务，把 `POST /v1/systemone` 契约带给
+任何本机工具。**不提供对外服务**——它监听 127.0.0.1，只服务你自己机器上的
+agent 与工作流。
+
+## 定位
+
+- 决策模型不是聊天模型：本服务**刻意不实现** chat / Ollama / OpenAI-compat。
+- 推理内核为**纯 torch 自写前向**（运行时零 laya/transformers 依赖），权重直读
+  发布的 hf_repo safetensors；与官方推理栈做过数值对齐（见 `dev_tools/verify_model.py`，
+  5 组跨题型用例，logits 最大偏差 0）。
+- 支持 noul（是否）/ choice（多选一）/ score（2–10 打分）三种题型，一次请求最多
+  64 题、choice 最多 255 选项、上下文预算 8192 token。
+
+## Install
+
+```bash
+git clone https://github.com/Phocinae/phocinae-server.git
+cd phocinae-server
+python -m venv .venv && . .venv/bin/activate
+pip install fastapi uvicorn torch
+```
+
+模型权重：设置 `PHOC_MODEL_DIR` 指向斑海豹 hf_repo 目录（含
+`model.safetensors`、`encoder/config.json`、`rl_agent_config.json`、`tokenizer/`）。
+
+## 运行
+
+```bash
+PHOC_MODEL_DIR=/path/to/hf_repo python -m phocinae.main
+# 默认 http://127.0.0.1:8155 ，/docs 有交互式文档
+```
+
+环境变量：
+
+| 变量 | 默认 | 说明 |
+|---|---|---|
+| `PHOC_MODEL_DIR` | (必设) | hf_repo 目录 |
+| `PHOC_HOST` / `PHOC_PORT` | 127.0.0.1 / 8155 | 监听地址 |
+| `PHOC_DEVICE` | auto | auto / cuda / cpu |
+| `PHOC_PERM_AVG` | 0 | 1 = 对 choice 题做 4 种选项排列平均（压翻转） |
+| `PHOC_BEARER_TOKEN` | 空 | 设置后 /v1/* 要求 `Authorization: Bearer <token>` |
+| `PHOC_THREADS` | — | CPU 模式 torch 线程数 |
+| `PHOC_MAX_BODY` | 2MiB | 请求体上限（超限 413） |
+| `PHOC_EXTENSIONS` | 1 | 0 = 不返回 answer_confidence/action/routing 扩展键 |
+
+## API
+
+```
+GET  /health                存活 + 模型/设备信息（无需鉴权）
+GET  /v1/models             模型目录
+POST /v1/systemone          单请求判定
+POST /v1/systemone/permute  同左，choice 题选项排列平均
+POST /v1/systemone/batch    批量（≤64 个请求）
+```
+
+请求：
+
+```json
+{"model": "Phocinae-Largha-150M-v1",
+ "state": "agent 刚刚执行了 rm -rf /var/log/app",
+ "questions": [
+   {"id": "ok", "type": "noul", "threshold": 0.65},
+   {"id": "act", "type": "choice",
+    "options": ["allow", "ask", "deny"]},
+   {"id": "risk", "type": "score"}]}
+```
+
+响应：`answers` 里 noul=布尔、choice=选项下标、score=2–10 整数；`usage` 给 token
+计数；`answer_confidence`（温度标定的 top 概率）、`action.act_probability`、
+`routing`（设备/后端/排列平均开关）为扩展键，可关闭。
+
+## 性能（实测，本机 RTX 5090 / 20 核 CPU）
+
+| 路径 | p50 |
+|---|---|
+| GPU fp16 + torch.compile（整请求） | **1.75 ms** |
+| CPU fp32 8 线程（整请求） | 35.97 ms |
+
+GPU 前向经 `torch.compile(reduce-overhead, dynamic)` 融合（eager 14.1ms → 编译后
+~2ms）；加载时预热常用 batch 形状，首问无形状抖动。
+
+## 安全声明
+
+- 默认**仅监听 127.0.0.1**，默认无鉴权——不要把它暴露到局域网/公网；需要时用
+  `PHOC_BEARER_TOKEN` 加 bearer。
+- 判定结果来自 150M 决策模型，**不是安全审计的完备防护**；用于审批门时必须配合
+  确定性规则层（见 phocinae-guard 的 L0 表），并把灰区交给人工。
+- 服务崩溃时应由调用方 fail-closed（本服务不保证任何高可用）。
+
+## 兼容性
+
+- typesafe-sdk-python：把 base_url 指向本服务即可接入（字段按官方形状对齐）。
+- 降级链：无 GPU 自动回退 CPU（fp32）；GPU 显存不足时 torch 侧回退由上层处理。
+
+## Router 组件（escalate 升级门 + 工具路由）
+
+`phocinae/router.py` 提供两个 harness 决策组件，共用本服务推理后端，支持两种
+接入模式：**进程内 Engine**（零网络、最低延迟）与 **HTTP**（指向
+`http://127.0.0.1:8155`）。
+
+```python
+from phocinae.engine import Engine
+from phocinae.router import Router
+
+router = Router(engine=Engine(MODEL_DIR, device="cpu"))     # 进程内
+router = Router(base_url="http://127.0.0.1:8155")           # HTTP（另需 requests）
+```
+
+### escalate 升级门（E1，τ=0.6 冻结口径）
+
+本地单遍判定一次，`answer_confidence`（温度标定 top 概率，已与 laya canonical
+对齐）低于 τ 的决策升级外部模型/人工：
+
+```python
+res = router.escalate(
+    "The agent was asked to restart the web server. It checked the logs and restarted nginx.",
+    [{"id": "ok", "type": "noul"}])
+# 单题 → {"decision": bool/int, "confidence": 0.786, "escalate": False,
+#         "tau": 0.6, "gate": "E1-tau0.6", "question_id": "ok", "usage": {...}}
+# 多题 → {"decision": {id:值}, "confidence": {id:float}, "escalate": bool,
+#         "escalated_ids": [...], "per_question": [...], "tau", "gate", "usage"}
+```
+
+**口径（2026-10-07 冻结，实测）**：τ=0.6；E1 typed en 400 例本地 acc 0.789 →
+组合 acc 0.7948（+0.6pp，噪声内无劣化）；大模型调用占比 100%→18%，**省费 82%**；
+门控决策 perm 翻转 8.87% ≤ 13% 基线。证据：
+`09_常态探索/release_prep_20261007/08_发布材料_20261007/04_省费评估_20261007.md`。
+τ 缺省为冻结值 0.6，允许按调用覆盖（`tau=` 参数）——**新域须重扫 τ**，勿直接
+沿用。HTTP 模式要求服务端扩展键开启（`PHOC_EXTENSIONS=1`，默认开启）。
+
+### 工具路由（k≤10）
+
+state + ≤10 个工具描述 → choice 题（options=工具名，附一行描述时渲染为
+`name: description`，banking77 渲染先例）→ 4 排列平均（
+`/v1/systemone/permute` 口径，压选项顺序翻转）：
+
+```python
+res = router.route_tool(
+    "The user wants to find the latest news about the product launch.",
+    [{"name": "web_search", "description": "search the public web"},
+     {"name": "read_file", "description": "read a local file"},
+     {"name": "run_command", "description": "run a shell command"},
+     {"name": "list_files", "description": "list directory files"},
+     {"name": "fetch_url", "description": "fetch a URL"},
+     {"name": "ask_user", "description": "ask the human user"}])
+# → {"tool": "web_search", "tool_index": 0, "confidence": 0.5561,
+#    "per_option_scores": {"web_search": 0.5561, ...},  # 排列平均后的标定概率
+#    "per_question": [{"id":"tool","tool":...,"confidence":...,"escalate":...}],
+#    "escalate": bool, "tau": 0.6, "gate": "E1-tau0.6", "usage", "routing"}
+```
+
+`tools` 支持三种形态：`["name", ...]`、`[{"name", "description"}]`、
+`{"name": "description"}`。工具名须唯一且非空；k>10 直接报错（JevBench
+tool_selection 域界）。口径：k≤10 菜单 JevBench 12/12=1.0（本地原生零升级）；
+大菜单（k>10）不接终选，走层级路由或外部分诊。
+
+### 服务端配合
+
+工具路由的 `per_option_scores` 需要本服务新增的 `option_scores` 扩展键（逐选项
+标定概率向量：choice=请求选项序、noul=[p(false), p(true)]、score=2–10 各档；
+perm 时已按 4 排列平均）。引擎模式与 HTTP 模式输出同构。
+
+### 测试
+
+```bash
+CUDA_VISIBLE_DEVICES='' PHOC_DEVICE=cpu \
+  /home/hermes/decision-model/.venv/bin/python tests/test_router.py
+# 可选 HTTP 对拍（额外拉起一个本机 CPU server 实例）：加 PHOC_ROUTER_HTTP_TEST=1
+```
+
+覆盖：escalate 门低/高置信样本各若干条（本机 CPU 实测标定 conf）、多题聚合、
+τ 覆盖、6 工具 ×4 场景路由、确定性（两次运行逐位一致）、参数校验（k>10/重名/
+空表）、HTTP 与引擎模式对拍（决策一致、conf 偏差 ≤0.005）。
+
+## License
+
+Apache-2.0（见 LICENSE）。模型权重另见其发布的模型卡。

phocinae_server-0.1.0/README.md

@@ -0,0 +1,180 @@
+# phocinae-server
+
+斑海豹（Phocinae-Largha-150M）本机决策服务：把 1.5 亿参数的决策模型打包成
+一行 `pip install` 级别依赖的 FastAPI 服务，把 `POST /v1/systemone` 契约带给
+任何本机工具。**不提供对外服务**——它监听 127.0.0.1，只服务你自己机器上的
+agent 与工作流。
+
+## 定位
+
+- 决策模型不是聊天模型：本服务**刻意不实现** chat / Ollama / OpenAI-compat。
+- 推理内核为**纯 torch 自写前向**（运行时零 laya/transformers 依赖），权重直读
+  发布的 hf_repo safetensors；与官方推理栈做过数值对齐（见 `dev_tools/verify_model.py`，
+  5 组跨题型用例，logits 最大偏差 0）。
+- 支持 noul（是否）/ choice（多选一）/ score（2–10 打分）三种题型，一次请求最多
+  64 题、choice 最多 255 选项、上下文预算 8192 token。
+
+## Install
+
+```bash
+git clone https://github.com/Phocinae/phocinae-server.git
+cd phocinae-server
+python -m venv .venv && . .venv/bin/activate
+pip install fastapi uvicorn torch
+```
+
+模型权重：设置 `PHOC_MODEL_DIR` 指向斑海豹 hf_repo 目录（含
+`model.safetensors`、`encoder/config.json`、`rl_agent_config.json`、`tokenizer/`）。
+
+## 运行
+
+```bash
+PHOC_MODEL_DIR=/path/to/hf_repo python -m phocinae.main
+# 默认 http://127.0.0.1:8155 ，/docs 有交互式文档
+```
+
+环境变量：
+
+| 变量 | 默认 | 说明 |
+|---|---|---|
+| `PHOC_MODEL_DIR` | (必设) | hf_repo 目录 |
+| `PHOC_HOST` / `PHOC_PORT` | 127.0.0.1 / 8155 | 监听地址 |
+| `PHOC_DEVICE` | auto | auto / cuda / cpu |
+| `PHOC_PERM_AVG` | 0 | 1 = 对 choice 题做 4 种选项排列平均（压翻转） |
+| `PHOC_BEARER_TOKEN` | 空 | 设置后 /v1/* 要求 `Authorization: Bearer <token>` |
+| `PHOC_THREADS` | — | CPU 模式 torch 线程数 |
+| `PHOC_MAX_BODY` | 2MiB | 请求体上限（超限 413） |
+| `PHOC_EXTENSIONS` | 1 | 0 = 不返回 answer_confidence/action/routing 扩展键 |
+
+## API
+
+```
+GET  /health                存活 + 模型/设备信息（无需鉴权）
+GET  /v1/models             模型目录
+POST /v1/systemone          单请求判定
+POST /v1/systemone/permute  同左，choice 题选项排列平均
+POST /v1/systemone/batch    批量（≤64 个请求）
+```
+
+请求：
+
+```json
+{"model": "Phocinae-Largha-150M-v1",
+ "state": "agent 刚刚执行了 rm -rf /var/log/app",
+ "questions": [
+   {"id": "ok", "type": "noul", "threshold": 0.65},
+   {"id": "act", "type": "choice",
+    "options": ["allow", "ask", "deny"]},
+   {"id": "risk", "type": "score"}]}
+```
+
+响应：`answers` 里 noul=布尔、choice=选项下标、score=2–10 整数；`usage` 给 token
+计数；`answer_confidence`（温度标定的 top 概率）、`action.act_probability`、
+`routing`（设备/后端/排列平均开关）为扩展键，可关闭。
+
+## 性能（实测，本机 RTX 5090 / 20 核 CPU）
+
+| 路径 | p50 |
+|---|---|
+| GPU fp16 + torch.compile（整请求） | **1.75 ms** |
+| CPU fp32 8 线程（整请求） | 35.97 ms |
+
+GPU 前向经 `torch.compile(reduce-overhead, dynamic)` 融合（eager 14.1ms → 编译后
+~2ms）；加载时预热常用 batch 形状，首问无形状抖动。
+
+## 安全声明
+
+- 默认**仅监听 127.0.0.1**，默认无鉴权——不要把它暴露到局域网/公网；需要时用
+  `PHOC_BEARER_TOKEN` 加 bearer。
+- 判定结果来自 150M 决策模型，**不是安全审计的完备防护**；用于审批门时必须配合
+  确定性规则层（见 phocinae-guard 的 L0 表），并把灰区交给人工。
+- 服务崩溃时应由调用方 fail-closed（本服务不保证任何高可用）。
+
+## 兼容性
+
+- typesafe-sdk-python：把 base_url 指向本服务即可接入（字段按官方形状对齐）。
+- 降级链：无 GPU 自动回退 CPU（fp32）；GPU 显存不足时 torch 侧回退由上层处理。
+
+## Router 组件（escalate 升级门 + 工具路由）
+
+`phocinae/router.py` 提供两个 harness 决策组件，共用本服务推理后端，支持两种
+接入模式：**进程内 Engine**（零网络、最低延迟）与 **HTTP**（指向
+`http://127.0.0.1:8155`）。
+
+```python
+from phocinae.engine import Engine
+from phocinae.router import Router
+
+router = Router(engine=Engine(MODEL_DIR, device="cpu"))     # 进程内
+router = Router(base_url="http://127.0.0.1:8155")           # HTTP（另需 requests）
+```
+
+### escalate 升级门（E1，τ=0.6 冻结口径）
+
+本地单遍判定一次，`answer_confidence`（温度标定 top 概率，已与 laya canonical
+对齐）低于 τ 的决策升级外部模型/人工：
+
+```python
+res = router.escalate(
+    "The agent was asked to restart the web server. It checked the logs and restarted nginx.",
+    [{"id": "ok", "type": "noul"}])
+# 单题 → {"decision": bool/int, "confidence": 0.786, "escalate": False,
+#         "tau": 0.6, "gate": "E1-tau0.6", "question_id": "ok", "usage": {...}}
+# 多题 → {"decision": {id:值}, "confidence": {id:float}, "escalate": bool,
+#         "escalated_ids": [...], "per_question": [...], "tau", "gate", "usage"}
+```
+
+**口径（2026-10-07 冻结，实测）**：τ=0.6；E1 typed en 400 例本地 acc 0.789 →
+组合 acc 0.7948（+0.6pp，噪声内无劣化）；大模型调用占比 100%→18%，**省费 82%**；
+门控决策 perm 翻转 8.87% ≤ 13% 基线。证据：
+`09_常态探索/release_prep_20261007/08_发布材料_20261007/04_省费评估_20261007.md`。
+τ 缺省为冻结值 0.6，允许按调用覆盖（`tau=` 参数）——**新域须重扫 τ**，勿直接
+沿用。HTTP 模式要求服务端扩展键开启（`PHOC_EXTENSIONS=1`，默认开启）。
+
+### 工具路由（k≤10）
+
+state + ≤10 个工具描述 → choice 题（options=工具名，附一行描述时渲染为
+`name: description`，banking77 渲染先例）→ 4 排列平均（
+`/v1/systemone/permute` 口径，压选项顺序翻转）：
+
+```python
+res = router.route_tool(
+    "The user wants to find the latest news about the product launch.",
+    [{"name": "web_search", "description": "search the public web"},
+     {"name": "read_file", "description": "read a local file"},
+     {"name": "run_command", "description": "run a shell command"},
+     {"name": "list_files", "description": "list directory files"},
+     {"name": "fetch_url", "description": "fetch a URL"},
+     {"name": "ask_user", "description": "ask the human user"}])
+# → {"tool": "web_search", "tool_index": 0, "confidence": 0.5561,
+#    "per_option_scores": {"web_search": 0.5561, ...},  # 排列平均后的标定概率
+#    "per_question": [{"id":"tool","tool":...,"confidence":...,"escalate":...}],
+#    "escalate": bool, "tau": 0.6, "gate": "E1-tau0.6", "usage", "routing"}
+```
+
+`tools` 支持三种形态：`["name", ...]`、`[{"name", "description"}]`、
+`{"name": "description"}`。工具名须唯一且非空；k>10 直接报错（JevBench
+tool_selection 域界）。口径：k≤10 菜单 JevBench 12/12=1.0（本地原生零升级）；
+大菜单（k>10）不接终选，走层级路由或外部分诊。
+
+### 服务端配合
+
+工具路由的 `per_option_scores` 需要本服务新增的 `option_scores` 扩展键（逐选项
+标定概率向量：choice=请求选项序、noul=[p(false), p(true)]、score=2–10 各档；
+perm 时已按 4 排列平均）。引擎模式与 HTTP 模式输出同构。
+
+### 测试
+
+```bash
+CUDA_VISIBLE_DEVICES='' PHOC_DEVICE=cpu \
+  /home/hermes/decision-model/.venv/bin/python tests/test_router.py
+# 可选 HTTP 对拍（额外拉起一个本机 CPU server 实例）：加 PHOC_ROUTER_HTTP_TEST=1
+```
+
+覆盖：escalate 门低/高置信样本各若干条（本机 CPU 实测标定 conf）、多题聚合、
+τ 覆盖、6 工具 ×4 场景路由、确定性（两次运行逐位一致）、参数校验（k>10/重名/
+空表）、HTTP 与引擎模式对拍（决策一致、conf 偏差 ≤0.005）。
+
+## License
+
+Apache-2.0（见 LICENSE）。模型权重另见其发布的模型卡。

phocinae_server-0.1.0/phocinae/__init__.py

@@ -0,0 +1,8 @@
+"""phocinae-server: self-contained typed-decisions (system-one) inference server.
+
+Zero runtime dependencies beyond fastapi + uvicorn + torch:
+tokenizer, safetensors loading, mmBERT encoder and decision head are all
+implemented here in pure python/torch (no laya, no transformers, no tokenizers).
+"""
+
+__version__ = "0.1.0"