feat: 完成三角色工作台及金租对公Skills全流程演示

backend/tools/skills_scanner.py

@@ -1,11 +1,64 @@
 """Skills Scanner — Scan skills/ directory and generate SKILLS_SNAPSHOT.md."""
 
-from pathlib import Path
 from hashlib import sha256
+from pathlib import Path
+import re
 
 import yaml
 
 
+MAX_SKILL_BYTES = 64 * 1024
+
+
+def parse_skill(content: str) -> dict:
+    """Validate the small, declarative SKILL.md format used by the workbench."""
+    if not isinstance(content, str) or not content.startswith("---") or len(content.encode("utf-8")) > MAX_SKILL_BYTES:
+        raise ValueError("SKILL.md 缺少 YAML 头或超过 64 KB")
+    parts = content.split("---", 2)
+    meta = yaml.safe_load(parts[1]) if len(parts) == 3 else None
+    if not isinstance(meta, dict) or not isinstance(meta.get("name"), str) or not meta["name"].strip():
+        raise ValueError("SKILL.md 必须声明 name")
+    if not isinstance(meta.get("description"), str) or not meta["description"].strip():
+        raise ValueError("SKILL.md 必须声明 description")
+    details = meta.get("metadata", {})
+    if not isinstance(details, dict):
+        raise ValueError("SKILL.md 的 metadata 必须是对象")
+    for field in ("inputs", "audience"):
+        value = meta.get(field, details.get(field))
+        if value is not None and (not isinstance(value, list) or not all(isinstance(item, str) for item in value)):
+            raise ValueError(f"SKILL.md 的 {field} 必须是文本列表")
+    return meta
+
+
+def skill_attributes(meta: dict) -> dict:
+    return {**meta, **meta.get("metadata", {})}
+
+
+def skill_slug(meta: dict) -> str:
+    slug = re.sub(r"[^a-z0-9-]+", "-", meta["name"].strip().lower().replace("_", "-")).strip("-")
+    if not slug or len(slug) > 64:
+        raise ValueError("Skill 名称需使用 1–64 位英文小写字母、数字或连字符")
+    return slug
+
+
+def install_skill(base_dir: Path, content: str) -> dict:
+    """Install one uploaded SKILL.md under the managed imported-skills directory."""
+    meta = parse_skill(content)
+    slug = skill_slug(meta)
+    root = (base_dir / "skills").resolve()
+    target = (root / "imported" / slug / "SKILL.md").resolve()
+    if not target.is_relative_to(root):
+        raise ValueError("Skill 目录不在允许范围内")
+    if target.exists():
+        raise FileExistsError("同名 Skill 已存在，请先归档旧版本或修改 name")
+    target.parent.mkdir(parents=True, exist_ok=False)
+    temporary = target.with_suffix(".tmp")
+    temporary.write_text(content, encoding="utf-8")
+    temporary.replace(target)
+    legacy_id = target.relative_to(root).as_posix()
+    return {"id": legacy_id, "legacy_id": legacy_id, "name": meta["name"], "description": meta["description"], "meta": skill_attributes(meta)}
+
+
 def discover_skills(base_dir: Path) -> list[dict]:
     """Read the native Skills catalog without changing its generated snapshot."""
     root = (base_dir / "skills").resolve()
@@ -16,21 +69,19 @@ def discover_skills(base_dir: Path) -> list[dict]:
             continue
         try:
             content = file.read_text(encoding="utf-8")
-            if not content.startswith("---"):
-                continue
-            parts = content.split("---", 2)
-            meta = yaml.safe_load(parts[1]) if len(parts) == 3 else None
-            if not isinstance(meta, dict):
+            meta = parse_skill(content)
+            attributes = skill_attributes(meta)
+            if str(attributes.get("status", "published")) == "archived":
                 continue
             legacy_id = file.relative_to(root).as_posix()
             entries.append({
                 "id": legacy_id,
-                "component_id": str(meta.get("id", legacy_id)),
+                "component_id": str(attributes.get("id", legacy_id)),
                 "legacy_id": legacy_id,
-                "kind": str(meta.get("kind", "skill")),
-                "version": str(meta.get("version", "legacy")),
-                "schema_version": int(meta.get("schema_version", 1)),
-                "status": str(meta.get("status", "published")),
+                "kind": str(attributes.get("kind", "skill")),
+                "version": str(attributes.get("version", "legacy")),
+                "schema_version": int(attributes.get("schema_version", 1)),
+                "status": str(attributes.get("status", "published")),
                 "name": str(meta.get("name", file.parent.name)),
                 "description": str(meta.get("description", "")),
                 "content_hash": sha256(content.encode("utf-8")).hexdigest(),