Guide flexible prompt poses through MCP

CHANGELOG.md

@@ -2,6 +2,9 @@
 
 ## v1.63.0 — unreleased
 
+- AI 写出图提示词时，AMB 的工具说明会建议：姿势可以选时优先选模型容易画清楚的动作，避免不必要的复杂手势；你指定的姿势和提供的文本照你的要求处理。复杂姿势可以用参考姿势或分步修复，并检查结果。
+- When an AI composes an image prompt, AMB's tool descriptions suggest a clear pose suited to the model when the pose is flexible. Explicit poses and supplied text are preserved; pose references or staged repair can help with difficult poses.
+
 - 跑图前的尺寸提醒放宽一档：6GB 显卡采样到约 420 万像素才提醒（以前 210 万），12GB 约 840 万。今天在 6GB 的笔记本显卡上实测，400 万像素一次约 100 秒，并不吃紧；以前的门槛会让 AI 动不动就问你要不要停。解码大图的提醒不变。
 - The size warning before a run is less eager: on a 6 GB card from about 4.2 megapixels (2.1 before), 8.4 on 12 GB. A 6 GB laptop card sampled 4 megapixels in about 100 s today; the old line had the AI ask you to stop runs that were fine. The warning for decoding a large picture is as before.
 

api/mcp_actions.py

@@ -257,6 +257,10 @@ def speak(ctx, character, text, seed=None, language="auto"):
 
 _NODE = {"type": "integer", "description": "A node id."}
 _MODEL = {"type": "string", "description": "model_id from search_models."}
+_PROMPT = {"type": "string", "description": "When composing a prompt, if the pose is flexible, prefer a clear, "
+           "natural pose suited to the model; avoid unnecessary interlocked hands or complex contact. Preserve "
+           "explicit user requests and supplied text. For difficult requested poses, consider pose references "
+           "or staged repair and inspect the result."}
 
 
 def _action(name, title, description, run, properties=None, required=(), destructive=False, open_world=False,
@@ -278,7 +282,7 @@ ACTIONS = dict([
     _action("set_prompt", "Write a prompt",
             "Writes text into the workflow's positive or negative prompt box (or the one on node_id). "
             "mode=append adds to what is there. One Ctrl+Z step.",
-            set_prompt, {"prompt": {"type": "string"}, "role": {"type": "string", "enum": ["positive", "negative"], "default": "positive"},
+            set_prompt, {"prompt": _PROMPT, "role": {"type": "string", "enum": ["positive", "negative"], "default": "positive"},
                          "node_id": _NODE, "mode": {"type": "string", "enum": ["replace", "append"], "default": "replace"}},
             ["prompt"], destructive=True),
     _action("set_model", "Swap a node's model",
@@ -332,7 +336,7 @@ ACTIONS = dict([
                            "wait_seconds": {"type": "number", "minimum": 0, "maximum": 55, "default": 55,
                                             "description": "Seconds to wait for it to end."},
                            "workflow": {"type": "string", "description": "A name from list_workflows."},
-                           "prompt": {"type": "string"}, "negative": {"type": "string"},
+                           "prompt": _PROMPT, "negative": {"type": "string"},
                            "seed": {"type": "integer", "minimum": 0},
                            "width": {"type": "integer", "minimum": 16, "maximum": 16384},
                            "height": {"type": "integer", "minimum": 16, "maximum": 16384},

docs/architecture/backend.md

@@ -252,6 +252,10 @@ descriptions. Its picks go to `/anomalous/scan_all` as `targets`, like a list pa
 
 ## MCP endpoint
 
+The `prompt` schemas for `set_prompt` and `run_workflow` share concise pose guidance for
+agents composing prompts: choose a model-appropriate pose when flexible and preserve explicit
+requests and supplied text. This is advice in the MCP tool description; AMB passes text through.
+
 `track_run(prompt_ids, task, wait_seconds)` attaches existing submissions from this server to
 an open task without running anything. IDs must exist in its queue/history; each is owned by
 one recorded run. Retrying an identical batch returns that run, not the task's newest run.

docs/guides/mcp.md

@@ -103,7 +103,7 @@ claude mcp add --transport http anomalous http://127.0.0.1:8188/anomalous/mcp
 | find_nodes | 搜索最多五个节点摘要；需要具体参数和选项时，用 `types=[节点类型]` 读取 |
 | add_node / connect_nodes | 添加节点并按唯一类型自动连线，或把节点插进已有路径；见下方例子 |
 | add_nodes | 一次加一组节点：按 ComfyUI 的 API 格式写，组里的节点互相连，也能连画布上已有的节点（写 `"#12"`，输出写编号或名字），排成一块放在空位上，整组算一步撤销 |
-| set_prompt | 改写或追加正向/反向提示词 |
+| set_prompt | 改写或追加正向/反向提示词。AI 自己写提示词时，姿势可以选择的话会优先考虑模型容易画清楚的动作；你指定的姿势和提供的文本仍按你的要求处理。`run_workflow` 的提示词也有同样的建议 |
 | set_model | 换某个加载器节点里的模型 |
 | add_lora | 插入一个 LoRA，默认接在主模型加载器和已有的 LoRA 后面。LoRA 文件里练过文本编码器的（SD1.5、SDXL、Illustrious 的大多是），CLIP 线也接上：从 checkpoint 接，或者从 UNet 工作流里单独的 CLIP 加载器接，正反向提示词一起改接到它后面；没练过的（多数 Flux、Anima 的 LoRA）只接模型线。LoRA 和主模型的底模对不上时（SD 1.5 的 LoRA 用在 SDXL 上），回复里会提醒 AI 并告诉它按底模去找；换模型（set_model）也一样 |
 | place_combo | 把一个搭配作为一组新节点摆到画布上，原有节点不动 |