feat: multimodal image clients, workspaces, Tushare skill, test fixes

- Replace monolithic image_message_client with HTTP/OCR/legacy modules; tighten OpenAI transport + tool schemas for multimodal downgrade to OCR specialist path.
- Add image/stock workspace prompts (META/SOUL/ROLE_SYSTEM); register experts; tweak specialist agent/direct loop/query_image_attachment.
- Add bundled runtime/skills/tushare-finance (references, api_client, SKILL metadata).
- Document OCR-related env vars; admin chat tweaks; README; weixin_install Ensure-OfficialPluginRuntimeDeps helper.
- Tests: multimodal downgrade + OCR coverage, strict tool pairing in attachment replay guard, workspace contract skips _internal/_system dirs, router/trace/prompt guards.

Co-authored-by: Cursor <cursoragent@cursor.com>

runtime/agents/factory.py

@@ -275,6 +275,11 @@ def build_ops_agent(
     )
 
 
+# `default_registry` treats empty allow_tags + empty allow_tools as "no filter". Use an impossible
+# tool name so the image specialist gets an empty tool surface (vision-only turns).
+_IMAGE_SPECIALIST_TOOL_ALLOWLIST: tuple[str, ...] = ("__oclaw_image_specialist_no_tools__",)
+
+
 def build_gateway_executor(
     store: SqliteStore,
     *,
@@ -321,14 +326,17 @@ def build_gateway_executor(
             path_policy_tenant_id=path_policy_tenant_id,
             path_policy_user_id=path_policy_user_id,
         )
-    tools = default_registry(
-        expert=expert_name_for_specialist(prof.name),
-        specialist=prof.name,
-        policy_session_id=policy_session_id,
-        path_policy_tenant_id=path_policy_tenant_id,
-        path_policy_user_id=path_policy_user_id,
-        store=store,
-    )
+    reg_kw: dict[str, Any] = {
+        "expert": expert_name_for_specialist(prof.name),
+        "specialist": prof.name,
+        "policy_session_id": policy_session_id,
+        "path_policy_tenant_id": path_policy_tenant_id,
+        "path_policy_user_id": path_policy_user_id,
+        "store": store,
+    }
+    if prof.name == "image":
+        reg_kw["allow_tools"] = list(_IMAGE_SPECIALIST_TOOL_ALLOWLIST)
+    tools = default_registry(**reg_kw)
     return Agent(
         store=store,
         tools=tools,

runtime/agents/specialist_agent.py

@@ -15,7 +15,7 @@ from oclaw.runtime.chat.agent import GenerationInterrupted
 from oclaw.runtime.agents.network_ops_agent import NetworkOpsAgent
 from oclaw.platform.persistence.sqlite_store import SqliteStore
 from oclaw.platform.files.attachment_assets import AttachmentAssetStore, attachment_id_to_data_url
-from oclaw.platform.llm.image_message_client import send_image_messages
+from oclaw.platform.llm.image_ocr_client import send_ocr_image_messages
 from oclaw.runtime.tools import default_registry
 from oclaw.runtime.agents.specialists import expert_name_for_specialist
 
@@ -267,25 +267,22 @@ class SpecialistAgentRunner:
                 else:
                     _, chosen_model, _ = self._resolve_profile_and_model(step.specialist)
                     model_name = str(
-                        os.getenv("AIA_IMAGE_MODEL")
-                        or getattr(chosen_model, "model", None)
-                        or ""
+                        os.getenv("AIA_OCR_MODEL") or getattr(chosen_model, "model", None) or ""
                     ).strip() or None
                     api_key = str(getattr(chosen_model, "api_key", "") or "").strip() or None
                     base_url = str(getattr(chosen_model, "base_url", "") or "").strip() or None
-                    dashscope_api_key = api_key if base_url and "dashscope.aliyuncs.com" in base_url.lower() else None
-                    dashscope_base_http_api_url = None
-                    if base_url and "dashscope.aliyuncs.com" in base_url.lower():
-                        # Normalize compatible-mode/v1 to native /api/v1 for DashScope SDK.
-                        dashscope_base_http_api_url = str(base_url).replace("/compatible-mode/v1", "/api/v1")
-                    resp = send_image_messages(
+                    text_parts = [
+                        str(x).strip()
+                        for x in (step.objective, step.input_text, parent_task.user_text)
+                        if str(x or "").strip()
+                    ]
+                    user_text = "\n".join(text_parts) if text_parts else "请根据用户上传的图片作答：描述可见场景、物体与文字；不确定处请标明。"
+                    resp = send_ocr_image_messages(
                         images=selected_images,
-                        prompt=f"{step.objective}\n\n{step.input_text}\n\n{parent_task.user_text}",
+                        prompt=user_text,
                         model=model_name,
                         api_key=api_key,
                         base_url=base_url,
-                        dashscope_api_key=dashscope_api_key,
-                        dashscope_base_http_api_url=dashscope_base_http_api_url,
                     )
                     image_debug_schema = str(resp.get("debug_used_schema") or "").strip()
                     dbg = resp.get("debug_used_debug")

runtime/agents/specialists.py

@@ -37,10 +37,16 @@ SPECIALISTS: dict[SpecialistId, SpecialistConfig] = {
         expert_name="memory",
         default_tool_tags=None,
     ),
+    "image": SpecialistConfig(
+        specialist_id="image",
+        # Vision turns attach pixels in-message; gateway executor exposes no tools (see factory).
+        expert_name="image",
+        default_tool_tags=None,
+    ),
 }
 
 def discover_specialist_ids() -> tuple[SpecialistId, ...]:
-    rows = specialist_registry_snapshot(base_order=("generalist", "ops", "memory"))
+    rows = specialist_registry_snapshot(base_order=("generalist", "ops", "memory", "image"))
     return tuple(str(x.get("id") or "").strip().lower() for x in rows if str(x.get("id") or "").strip())