codewave-openclaw-installer 2.1.12 → 2.2.0

package/skills/ppt-master/references/image-generator.md

@@ -114,38 +114,12 @@ Full directive: "color palette: deep navy blue (#1E3A5F), light gray (#F8F9FA),
 
 ### Type Determination Flow
 
-0. **Architecture / topology / system overview → 🌟 Architecture Diagram (3.0) — HIGHEST PRIORITY, all modes**
 1. Full-page / large-area backdrop → **Background** (3.1)
 2. Real scenes / people / products → **Photography** (3.2)
 3. Flat / illustration / cartoon style → **Illustration** (3.3)
 4. Process / architecture / relationships → **Diagram** (3.4)
 5. Partial decoration / texture → **Decorative Pattern** (3.5)
 
-### 3.0 Architecture Diagram (🌟 Core Feature — All Modes)
-
-**Identifying characteristics**: Pages with titles containing "架构", "拓扑", "技术方案", "系统概览", "Architecture", "Topology", or any complex technical structure visualization.
-
-> ⚠️ **This is the skill's signature capability.** Architecture diagrams are ALWAYS generated via `image_gen.py`, regardless of Tier selection (Tier 1/2/3) or scenario (bid/free/template). They are never skipped.
-
-| Key Point | Description |
-|-----------|-------------|
-| Light color scheme | Prefer light/pastel tones matching PPT palette |
-| **No text in image** | Never include any text, labels, or words — text is rendered as SVG `<text>` elements on top |
-| Abstract/conceptual | Show interconnected modules, flowing data paths, layered structures |
-| Professional style | Enterprise-grade, clean, not cartoonish |
-| Semi-transparent usage | Typically used at 20-40% opacity as background layer behind SVG text |
-
-**Template**: `Light colored abstract {architecture_type} architecture concept, interconnected modules and components floating in soft gradient space, {theme_elements} connected by flowing light lines, clean minimal style, soft {primary_color} and light gray tones with subtle {accent_color} accents, no text no labels no words no numbers, pastel background, professional enterprise {domain} visualization, 16:9 aspect ratio`
-
-**Negative prompt**: `text, letters, words, numbers, labels, annotations, watermark, dark background, dark colors, busy patterns, cartoon, clipart`
-
-**SVG Integration**: Architecture diagram images are embedded as background layers:
-```xml
-<image href="../images/{arch_name}.png" x="0" y="80" width="1280" height="600" opacity="0.25"/>
-<!-- SVG text elements rendered on top -->
-<text x="640" y="200" font-size="28" font-weight="bold" text-anchor="middle">Architecture Title</text>
-```
-
 ### 3.1 Background
 
 **Identifying characteristics**: Full-page background for covers or chapter pages; must support text overlay
@@ -156,69 +130,10 @@ Full directive: "color palette: deep navy blue (#1E3A5F), light gray (#F8F9FA),
 | Reserve text area | `negative space in center for text overlay` |
 | Avoid strong subjects | Use abstract, gradient, geometric elements |
 | Low-contrast details | `subtle`, `soft`, `muted` |
-| **Light color scheme** | Prefer light/pastel tones; avoid dark backgrounds |
-| **No text in image** | Never include any text, letters, or words in generated images |
-
-**Template**: `Abstract {theme element} background, {style} style, light {primary color} to soft {secondary color} gradient, subtle {decorative elements}, clean negative space in center for text overlay, light color palette, pastel tones, {aspect ratio} aspect ratio, high resolution, professional presentation background`
-
-**Negative prompt**: `text, letters, words, numbers, watermark, faces, busy patterns, high contrast details, dark background, dark colors, black background`
-
-### 3.1.1 Content Page Background (Free Generation Mode Only)
-
-**When to use**: Only in free generation mode (non-bid, non-template), and only after **user explicitly confirms** they want AI backgrounds.
-
-**⚠️ Requires user confirmation**: Never auto-generate. Ask in Strategist phase and again before execution.
-
-**Cost optimization: Category-based reuse** (CRITICAL):
-Instead of generating one image per page, group pages by category and **share one background image per category**:
-
-| Category | Matching pages | Prompt keywords |
-|----------|---------------|-----------------|
-| cover | cover, ending | `corporate, modern business, company` |
-| chapter | chapter_* | `section divider, clean transition, minimal` |
-| company | company_*, brand_*, certifications, client_logos | `corporate headquarters, enterprise building` |
-| tech | platform_*, dev_*, function_*, source_*, ai_* | `technology, digital platform, software` |
-| product | asset_*, integration_*, permission_*, i18n_*, multi_* | `product modules, capability, ecosystem` |
-| cases | *_case, *_apps, cpq_*, operation_* | `enterprise partnership, factory, industry` |
-| service | tech_support, knowledge_*, sla_*, custom_*, strategic_*, ecosystem | `customer service, support center` |
-| delivery | org_*, implementation_*, core_team, pm_*, after_sales | `project management, teamwork, office` |
-
-**Result**: 55-page PPT → only **6~8 images** generated (85%+ cost saving).
-
-**Skip these pages** (already have rich visuals):
-- Pages generated by `arch_diagram.py` (architecture diagrams)
-- Table of contents page
-
-**Key rules**:
-| Rule | Description |
-|------|-------------|
-| **Color matching** | Must match PPT theme — primary blue → light blue-gray tones |
-| **Ultra-low opacity** | Embed at `opacity="0.15"` to `opacity="0.25"` |
-| **No text ever** | No text, numbers, labels in generated images |
-| **Soft & subtle** | `subtle`, `soft`, `muted`, `ethereal` — non-distracting |
-
-**Template**: `Light pastel abstract {category_keywords}, soft gradient, subtle {primary_color_name} tones, ethereal, professional business, no text no words, minimal clean, 16:9`
-
-**Negative prompt**: `text, letters, words, numbers, labels, annotations, watermark, dark background, dark colors, high contrast, busy patterns, faces, people`
-
-**SVG integration** (first child after `<svg>` opening tag):
-```xml
-<image href="../images/{category}_bg.png" x="0" y="0" width="1280" height="720" opacity="0.18"/>
-```
 
-**Workflow**:
-1. After all SVGs generated, classify pages into categories (see table above)
-2. **Check global resource library first**: `${SKILL_DIR}/resources/backgrounds/` — if a matching `{category}_bg.png` exists, copy it to project `images/` directly (zero API calls)
-3. Only for missing categories: report to user "Need to generate N new backgrounds (M reused from library). Proceed?"
-4. After confirmation, generate missing images; **save to BOTH project `images/` AND `resources/backgrounds/`** (so future projects can reuse)
-5. Inject `<image>` tag into all SVGs of that category
-6. Run `finalize_svg.py` (auto-embeds as base64)
+**Template**: `Abstract {theme element} background, {style} style, {primary color} to {secondary color} gradient, subtle {decorative elements}, clean negative space in center for text overlay, {aspect ratio} aspect ratio, high resolution, professional presentation background`
 
-**Global Resource Library**: `${SKILL_DIR}/resources/backgrounds/`
-- This is a **cross-project shared library**, not tied to any single project
-- Every generated background is archived here for future reuse
-- `index.json` tracks metadata (prompt, color tone, suitable scenarios)
-- Over time, the library grows — new projects need fewer (or zero) API calls
+**Negative prompt**: `text, letters, watermark, faces, busy patterns, high contrast details`
 
 ### 3.2 Photography
 
@@ -438,8 +353,7 @@ Abstract futuristic background with flowing digital waves...
 
 | Type | Recommended Negative Prompt |
 |------|---------------------------|
-| **🌟 Architecture diagram** | `text, letters, words, numbers, labels, annotations, watermark, dark background, dark colors, busy patterns, cartoon, clipart` |
-| Background | `text, letters, words, numbers, watermark, faces, busy patterns, high contrast details, dark background, dark colors` |
+| Background | `text, letters, watermark, faces, busy patterns, high contrast details` |
 | Photography | `watermark, text overlay, artificial, CGI, illustration, cartoon, distorted faces` |
 | Illustration | `realistic, photography, 3D render, complex textures, watermark` |
 | Diagram | `cluttered, messy, overlapping elements, dark background, realistic` |
@@ -447,7 +361,7 @@ Abstract futuristic background with flowing digital waves...
 
 ### Universal Negative Prompts
 
-- **Standard**: `text, watermark, signature, blurry, distorted, low quality, letters, words, numbers`
+- **Standard**: `text, watermark, signature, blurry, distorted, low quality`
 - **Extended** (people scenarios): `text, watermark, signature, blurry, low quality, distorted, extra fingers, mutated hands, poorly drawn face, bad anatomy, extra limbs, disfigured, deformed`
 
 ---
@@ -458,9 +372,8 @@ Abstract futuristic background with flowing digital waves...
 
 | Purpose | Default Inference |
 |---------|------------------|
-| **🌟 Architecture/topology page** | Light-colored abstract architecture concept, interconnected modules, no text, professional enterprise style |
-| Cover background | Light-toned abstract gradient background, no text, reserve central text area |
-| Chapter page background | Clean geometric pattern, light/pastel colors, no text |
+| Cover background | Abstract gradient background, reserve central text area |
+| Chapter page background | Clean geometric pattern, monochrome focus |
 | Team introduction page | Team collaboration scene illustration (flat style) |
 | Data display page | Clean geometric pattern or solid color background |
 | Product showcase | Product photography style, white or gradient background |

package/skills/ppt-master/references/strategist.md

@@ -144,17 +144,14 @@ Selection principle: Font size is based on **content density**, not design style
 
 ### h. Image Usage Confirmation
 
-> **图片生成能力已就绪**（AIGW Gemini 后端）。推荐选择 B（增强模式），在纯 SVG 基础上为封面、结尾等关键页面增加 AI 生成的背景图，提升视觉品质。
-
 | Option | Approach | Suitable Scenarios |
 |--------|----------|-------------------|
-| **A** | No images — pure SVG visual design | 追求极简/快速出稿；纯渐变+几何+图标 |
-| **B (推荐)** | SVG + AI 生成装饰图（封面背景、章节配图） | 专业演示场景，需要视觉冲击力。SVG 结构为主体，图片锦上添花 |
-| **C** | SVG + AI 生成全量配图 | 每个受益页面都配图，最高视觉品质，耗时最长 |
-| **D** | User-provided images | 已有图片素材 |
-| **E** | Placeholders | 图片后续补充 |
+| **A** | No images | Data reports, process documentation |
+| **B** | User-provided | Has existing image assets |
+| **C** | AI-generated | Custom illustrations, backgrounds needed |
+| **D** | Placeholders | Images to be added later |
 
-**When selection includes D**, you must run `python3 scripts/analyze_images.py <project_path>/images` before outputting the spec, and integrate scan results into the image resource list.
+**When selection includes B**, you must run `python3 scripts/analyze_images.py <project_path>/images` before outputting the spec, and integrate scan results into the image resource list.
 
 **When B/C/D is selected**, add an image resource list to the spec:
 

package/skills/ppt-master/scripts/README.md

@@ -28,7 +28,7 @@ python3 scripts/svg_to_pptx.py <project_path> -s final
 | Area | Primary scripts | Documentation |
 |------|-----------------|---------------|
 | Conversion | `pdf_to_md.py`, `doc_to_md.py`, `web_to_md.py`, `web_to_md.cjs` | [docs/conversion.md](./docs/conversion.md) |
-| Project management | `project_manager.py`, `bid_init.py`, `batch_validate.py`, `generate_examples_index.py`, `error_helper.py` | [docs/project.md](./docs/project.md) |
+| Project management | `project_manager.py`, `batch_validate.py`, `generate_examples_index.py`, `error_helper.py` | [docs/project.md](./docs/project.md) |
 | SVG pipeline | `finalize_svg.py`, `svg_to_pptx.py`, `total_md_split.py`, `svg_quality_checker.py` | [docs/svg-pipeline.md](./docs/svg-pipeline.md) |
 | Image tools | `image_gen.py`, `analyze_images.py`, `gemini_watermark_remover.py` | [docs/image.md](./docs/image.md) |
 | Troubleshooting | validation, preview, export, dependency issues | [docs/troubleshooting.md](./docs/troubleshooting.md) |

package/skills/ppt-master/scripts/finalize_svg.py

@@ -238,20 +238,6 @@ def finalize_project(
 
 def main() -> None:
     """Run the CLI entry point."""
-    try:
-        _main_impl()
-    except KeyboardInterrupt:
-        print("\nInterrupted by user.")
-        sys.exit(130)
-    except SystemExit:
-        raise
-    except Exception as e:
-        safe_print(f"[ERROR] Fatal error: {type(e).__name__}: {e}")
-        sys.exit(1)
-
-
-def _main_impl() -> None:
-    """Internal implementation of main CLI logic."""
     parser = argparse.ArgumentParser(
         description='PPT Master - SVG Post-processing Tool',
         formatter_class=argparse.RawDescriptionHelpFormatter,

package/skills/ppt-master/scripts/svg_quality_checker.py

@@ -89,18 +89,6 @@ class SVGQualityChecker:
             # 5. Check text wrapping methods
             self._check_text_elements(content, result)
 
-            # 6. Check text overflow (beyond viewBox boundaries)
-            self._check_text_overflow(content, result)
-
-            # 7. Check bottom red line presence
-            self._check_bottom_line(content, result)
-
-            # 8. Check file size (warn if too large)
-            file_size = svg_path.stat().st_size
-            if file_size > 5 * 1024 * 1024:  # 5MB
-                result['warnings'].append(f"Large SVG file: {file_size // 1024}KB (may slow down export)")
-            result['info']['file_size_kb'] = file_size // 1024
-
             # Determine pass/fail
             result['passed'] = len(result['errors']) == 0
 
@@ -218,13 +206,7 @@ class SVGQualityChecker:
         if re.search(r'<g[^>]*\sopacity\s*=', content_lower):
             result['errors'].append("Detected forbidden <g opacity> (set opacity on each child element individually)")
         if re.search(r'<image[^>]*\sopacity\s*=', content_lower):
-            # Allow opacity on background images (first <image> child of <svg>)
-            # Only flag non-background images with opacity
-            image_opacity_matches = re.findall(r'<image[^>]*\sopacity\s*=\s*"([^"]*)"[^>]*/?\s*>', content, re.IGNORECASE)
-            bg_pattern = re.search(r'<svg[^>]*>\s*\n?\s*<image[^>]*opacity', content, re.IGNORECASE)
-            non_bg_opacity_count = len(image_opacity_matches) - (1 if bg_pattern else 0)
-            if non_bg_opacity_count > 0:
-                result['errors'].append(f"Detected {non_bg_opacity_count} non-background <image opacity> (use overlay mask approach)")
+            result['errors'].append("Detected forbidden <image opacity> (use overlay mask approach)")
 
     def _check_fonts(self, content: str, result: Dict):
         """Check font usage"""
@@ -286,71 +268,6 @@ class SVGQualityChecker:
                 f"Detected {len(text_matches)} potentially overly long single-line text(s) (consider using tspan for wrapping)"
             )
 
-    def _check_text_overflow(self, content: str, result: Dict):
-        """Check if text elements extend beyond viewBox boundaries"""
-        # Get viewBox dimensions
-        viewbox_match = re.search(r'viewBox="0 0 (\d+) (\d+)"', content)
-        if not viewbox_match:
-            return
-        vb_width = int(viewbox_match.group(1))
-        vb_height = int(viewbox_match.group(2))
-
-        overflow_count = 0
-
-        # Check <text> x/y coordinates
-        for m in re.finditer(r'<text[^>]*\bx="([^"]*)"[^>]*\by="([^"]*)"', content):
-            try:
-                x = float(m.group(1))
-                y = float(m.group(2))
-                if x < -10 or x > vb_width + 10:
-                    overflow_count += 1
-                if y < -10 or y > vb_height + 50:  # allow small bottom overflow for descenders
-                    overflow_count += 1
-            except (ValueError, TypeError):
-                pass
-
-        # Check <tspan> y coordinates for vertical overflow
-        for m in re.finditer(r'<tspan[^>]*\by="([^"]*)"', content):
-            try:
-                y = float(m.group(1))
-                if y > vb_height + 20:
-                    overflow_count += 1
-            except (ValueError, TypeError):
-                pass
-
-        if overflow_count > 0:
-            result['errors'].append(
-                f"Detected {overflow_count} text element(s) potentially overflowing viewBox ({vb_width}x{vb_height})")
-
-    def _check_bottom_line(self, content: str, result: Dict):
-        """Check for bottom red line (required on every page)"""
-        # Look for a rect at y=716 with height=4 and red-ish fill
-        has_bottom_line = bool(re.search(
-            r'<rect[^>]*y="71[456]"[^>]*height="[3-6]"[^>]*fill="#[Cc]00000"',
-            content
-        )) or bool(re.search(
-            r'<rect[^>]*fill="#[Cc]00000"[^>]*y="71[456]"[^>]*height="[3-6]"',
-            content
-        ))
-        # Also check for line at bottom
-        if not has_bottom_line:
-            has_bottom_line = bool(re.search(
-                r'<line[^>]*y1="71[456]"[^>]*y2="71[456]"[^>]*stroke="#[Cc]00000"',
-                content
-            ))
-        if not has_bottom_line:
-            # More relaxed: any red rect near bottom
-            has_bottom_line = bool(re.search(
-                r'<rect[^>]*y="7[01]\d"[^>]*fill="#[Cc]00000"',
-                content
-            ))
-
-        if not has_bottom_line:
-            result['warnings'].append("Missing bottom red line (expected <rect y=\"716\" height=\"4\" fill=\"#C00000\"/>)")
-            result['info']['has_bottom_line'] = False
-        else:
-            result['info']['has_bottom_line'] = True
-
     def _categorize_issue(self, error_msg: str) -> str:
         """Categorize issue type"""
         if 'viewBox' in error_msg:
@@ -507,79 +424,24 @@ class SVGQualityChecker:
         print(f"\n[REPORT] Check report exported: {output_file}")
 
 
-def auto_fix_svg(svg_file: str) -> Dict:
-    """
-    Auto-fix common SVG issues.
-    Returns dict with fix counts.
-    """
-    path = Path(svg_file)
-    with open(path, 'r', encoding='utf-8') as f:
-        content = f.read()
-
-    original = content
-    fixes = []
-
-    # Fix 1: Replace rgba() with rgb() + opacity
-    rgba_matches = re.findall(r'rgba\(\s*(\d+)\s*,\s*(\d+)\s*,\s*(\d+)\s*,\s*([\d.]+)\s*\)', content)
-    for r, g, b, a in rgba_matches:
-        old = f'rgba({r},{g},{b},{a})'
-        # Also match with spaces
-        content = re.sub(
-            rf'rgba\(\s*{r}\s*,\s*{g}\s*,\s*{b}\s*,\s*{re.escape(a)}\s*\)',
-            f'rgb({r},{g},{b})',
-            content
-        )
-        fixes.append(f"rgba→rgb: rgba({r},{g},{b},{a})")
-
-    # Fix 2: Remove <style> blocks
-    if '<style' in content.lower():
-        content = re.sub(r'<style[^>]*>.*?</style>', '', content, flags=re.DOTALL | re.IGNORECASE)
-        fixes.append("Removed <style> block")
-
-    # Fix 3: Remove class attributes
-    if re.search(r'\bclass\s*=\s*"[^"]*"', content):
-        content = re.sub(r'\s*class\s*=\s*"[^"]*"', '', content)
-        fixes.append("Removed class attributes")
-
-    # Fix 4: Add missing bottom red line
-    has_bottom_line = bool(re.search(r'<rect[^>]*y="7[01]\d"[^>]*fill="#[Cc]00000"', content))
-    if not has_bottom_line:
-        # Insert before </svg>
-        bottom_line = '<rect x="0" y="716" width="1280" height="4" fill="#C00000"/>'
-        content = content.replace('</svg>', f'{bottom_line}\n</svg>')
-        fixes.append("Added bottom red line")
-
-    # Fix 5: Remove <g opacity="..."> by moving opacity to children
-    g_opacity_matches = re.findall(r'<g\s+opacity="([^"]+)">', content)
-    for op in g_opacity_matches:
-        # Simple removal of group opacity (conservative — just remove the opacity attr)
-        content = re.sub(r'<g\s+opacity="[^"]+"\s*>', '<g>', content, count=1)
-        fixes.append(f"Removed <g opacity=\"{op}\">")
-
-    if content != original:
-        with open(path, 'w', encoding='utf-8') as f:
-            f.write(content)
-
-    return {'file': path.name, 'fixes': fixes, 'changed': content != original}
-
-
 def main() -> None:
     """Run the CLI entry point."""
     if len(sys.argv) < 2:
         print("PPT Master - SVG Quality Check Tool\n")
         print("Usage:")
-        print("  python3 scripts/svg_quality_checker.py <svg_file_or_directory>")
-        print("  python3 scripts/svg_quality_checker.py <directory> --fix    # Auto-fix common issues")
-        print("  python3 scripts/svg_quality_checker.py <directory> --fix --check  # Fix then re-check")
+        print("  python3 scripts/svg_quality_checker.py <svg_file>")
+        print("  python3 scripts/svg_quality_checker.py <directory>")
         print("  python3 scripts/svg_quality_checker.py --all examples")
+        print("\nExamples:")
+        print("  python3 scripts/svg_quality_checker.py examples/project/svg_output/slide_01.svg")
+        print("  python3 scripts/svg_quality_checker.py examples/project/svg_output")
+        print("  python3 scripts/svg_quality_checker.py examples/project")
         sys.exit(0)
 
     checker = SVGQualityChecker()
-    do_fix = '--fix' in sys.argv
-    do_check_after_fix = '--check' in sys.argv
 
     # Parse arguments
-    target = [a for a in sys.argv[1:] if not a.startswith('--')][0] if any(not a.startswith('--') for a in sys.argv[1:]) else sys.argv[1]
+    target = sys.argv[1]
     expected_format = None
 
     if '--format' in sys.argv:
@@ -587,29 +449,6 @@ def main() -> None:
         if idx + 1 < len(sys.argv):
             expected_format = sys.argv[idx + 1]
 
-    # Auto-fix mode
-    if do_fix:
-        target_path = Path(target)
-        if target_path.is_file():
-            svg_files = [target_path]
-        else:
-            svg_output = target_path / 'svg_output' if (target_path / 'svg_output').exists() else target_path
-            svg_files = sorted(svg_output.glob('*.svg'))
-
-        print(f"\n[FIX] Auto-fixing {len(svg_files)} SVG file(s)...\n")
-        total_fixes = 0
-        for svg_file in svg_files:
-            result = auto_fix_svg(str(svg_file))
-            if result['changed']:
-                total_fixes += len(result['fixes'])
-                print(f"  [FIXED] {result['file']}: {', '.join(result['fixes'])}")
-        print(f"\n[FIX] Total fixes applied: {total_fixes}")
-
-        if not do_check_after_fix:
-            print("\n[TIP] Run with --fix --check to verify fixes")
-            sys.exit(0)
-        print("\n--- Re-checking after fixes ---\n")
-
     # Execute check
     if target == '--all':
         # Check all example projects

package/skills/ppt-master/scripts/svg_to_pptx/pptx_builder.py

@@ -351,16 +351,10 @@ def create_pptx_with_native_svg(
                     arcname = file_path.relative_to(extract_dir)
                     zf.write(file_path, arcname)
 
-        failed_count = len(svg_files) - success_count
         if verbose:
             print()
             print(f"[Done] Saved: {output_path}")
-            print(f"  Succeeded: {success_count}, Failed: {failed_count}")
-            if failed_count > 0:
-                failed_indices = [i + 1 for i in range(len(svg_files))
-                                  if i + 1 not in range(1, success_count + failed_count + 1)]
-                # List failed slides by checking which ones threw exceptions
-                print(f"  Note: {failed_count} slide(s) failed and were skipped.")
+            print(f"  Succeeded: {success_count}, Failed: {len(svg_files) - success_count}")
             if use_compat_mode and has_any_image:
                 print(f"  Mode: Office compatibility mode (supports all Office versions)")
                 if PNG_RENDERER == 'svglib' and renderer_hint:

package/skills/ppt-master/scripts/svg_to_pptx/pptx_cli.py

@@ -15,20 +15,6 @@ from .pptx_slide_xml import TRANSITIONS
 
 def main() -> None:
     """CLI entry point for the SVG to PPTX conversion tool."""
-    try:
-        _main_impl()
-    except KeyboardInterrupt:
-        print("\nInterrupted by user.")
-        sys.exit(130)
-    except SystemExit:
-        raise
-    except Exception as e:
-        print(f"Fatal error: {type(e).__name__}: {e}", file=sys.stderr)
-        sys.exit(1)
-
-
-def _main_impl() -> None:
-    """Internal implementation of main CLI logic."""
     transition_choices = (
         ['none'] + (list(TRANSITIONS.keys()) if TRANSITIONS
                     else ['fade', 'push', 'wipe', 'split', 'strips', 'cover', 'random'])

package/skills/risk-alert/SKILL.md

@@ -0,0 +1,110 @@
+# Risk Alert - 智能风险预警 Skill
+
+## 功能概述
+
+自动监控飞书群聊消息，智能识别风险并推送预警卡片。
+
+## 首次使用配置
+
+### 1. 配置飞书应用凭证
+
+需要用户提供飞书自建应用的 `App ID` 和 `App Secret`：
+
+```bash
+# 方式1：环境变量
+export FEISHU_APP_ID="your_app_id"
+export FEISHU_APP_SECRET="your_app_secret"
+
+# 方式2：配置文件（首次运行会自动创建）
+# 编辑 config.json 文件
+```
+
+### 2. 选择监控群聊
+
+首次运行时会自动获取机器人加入的所有群聊，并展示列表让用户确认：
+
+```
+📋 发现机器人已加入以下群聊：
+1. 项目群A (oc_xxx1)
+2. 项目群B (oc_xxx2)
+3. 测试群 (oc_xxx3)
+
+请选择要监控的群聊（输入序号，多个用逗号分隔，输入 all 选择全部）：
+```
+
+### 3. 选择推送目标
+
+```
+请选择风险预警推送目标：
+1. 推送到指定群聊
+2. 推送到指定用户
+3. 同时推送
+
+请输入选择（1/2/3）：
+```
+
+### 4. 保存配置
+
+配置会自动保存到 `config.json`，后续运行直接使用已保存的配置。
+
+## 使用方法
+
+### 手动执行
+
+```bash
+# 进入 skill 目录
+cd /Users/song/.openclaw/workspace/skills/risk-alert
+
+# 执行完整流程
+node scripts/risk-alert.js
+
+# 仅抓取消息
+node scripts/analyze-risk.js
+
+# 仅推送卡片（AI 分析完成后）
+node scripts/push-beautiful.js
+```
+
+### 通过触发器
+
+```bash
+# 发送触发指令给 AI
+./trigger-risk-alert.sh
+```
+
+## 工作流程
+
+```
+1. 抓取群聊消息 → raw-messages.json
+2. AI 智能分析 → risk-data.json  
+3. 推送预警卡片 → 飞书
+```
+
+## 配置项说明
+
+| 配置项 | 说明 | 必填 |
+|--------|------|------|
+| appId | 飞书应用 ID | 是 |
+| appSecret | 飞书应用 Secret | 是 |
+| monitoredChats | 监控的群聊列表 | 是 |
+| pushTarget.type | 推送类型: chat/user/both | 是 |
+| pushTarget.chatId | 推送群聊 ID | 条件必填 |
+| pushTarget.userId | 推送用户 ID | 条件必填 |
+| days | 首次拉取天数 | 否，默认7 |
+
+## 文件结构
+
+```
+risk-alert/
+├── SKILL.md              # 本文件
+├── config.json           # 用户配置（首次运行生成）
+├── trigger-risk-alert.sh # 触发脚本
+├── scheduler.sh          # 定时任务
+└── scripts/
+    ├── analyze-risk.js   # 消息抓取
+    ├── risk-alert.js     # 流程控制
+    ├── push-beautiful.js # 卡片推送
+    ├── config-manager.js # 配置管理
+    └── lib/
+        └── feishu-api.js # 飞书 API 封装
+```

package/skills/risk-alert/config.example.json

@@ -0,0 +1,21 @@
+{
+  "feishu": {
+    "appId": "YOUR_APP_ID_HERE",
+    "appSecret": "YOUR_APP_SECRET_HERE"
+  },
+  "monitoredChats": [
+    {
+      "id": "oc_xxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
+      "name": "群聊名称"
+    }
+  ],
+  "pushTarget": {
+    "type": "chat",
+    "chatId": "oc_xxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
+    "userId": ""
+  },
+  "settings": {
+    "days": 7,
+    "incrementMode": true
+  }
+}

package/skills/risk-alert/scheduler.sh

@@ -0,0 +1,74 @@
+#!/bin/bash
+# Risk Alert 定时推送脚本
+# 每20分钟执行一次
+
+# 设置 PATH（launchd 环境变量不同）
+export PATH="/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:$PATH"
+
+# 查找 node 路径
+NODE_PATH=$(which node)
+if [ -z "$NODE_PATH" ]; then
+    # 尝试常见路径
+    if [ -f "/usr/local/bin/node" ]; then
+        NODE_PATH="/usr/local/bin/node"
+    elif [ -f "/opt/homebrew/bin/node" ]; then
+        NODE_PATH="/opt/homebrew/bin/node"
+    elif [ -f "$HOME/.nvm/versions/node/v22.18.0/bin/node" ]; then
+        NODE_PATH="$HOME/.nvm/versions/node/v22.18.0/bin/node"
+    fi
+fi
+
+# 工作目录
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+cd "$SCRIPT_DIR/scripts"
+
+# 日志文件
+LOG_FILE="$SCRIPT_DIR/risk-alert.log"
+
+echo "[$(date '+%Y-%m-%d %H:%M:%S')] 开始执行 Risk Alert 定时任务" >> "$LOG_FILE"
+
+# Step 1: 全量抓取消息（删除状态文件）
+echo "[$(date '+%Y-%m-%d %H:%M:%S')] 步骤1: 抓取消息..." >> "$LOG_FILE"
+rm -f last-fetch-state.json
+$NODE_PATH analyze-risk.js >> "$LOG_FILE" 2>&1
+if [ $? -ne 0 ]; then
+    echo "[$(date '+%Y-%m-%d %H:%M:%S')] ❌ 消息抓取失败" >> "$LOG_FILE"
+    exit 1
+fi
+
+# 检查是否有新消息
+if [ -f "raw-messages.json" ]; then
+    NEW_COUNT=$($NODE_PATH -e "console.log(JSON.parse(require('fs').readFileSync('raw-messages.json')).newMessagesCount || 0)")
+    echo "[$(date '+%Y-%m-%d %H:%M:%S')] 新增消息: $NEW_COUNT 条" >> "$LOG_FILE"
+    
+    if [ "$NEW_COUNT" -eq 0 ]; then
+        echo "[$(date '+%Y-%m-%d %H:%M:%S')] ℹ️ 没有新消息，跳过分析" >> "$LOG_FILE"
+        # 仍然推送一次（显示无风险状态）
+        $NODE_PATH push-beautiful.js >> "$LOG_FILE" 2>&1
+        echo "[$(date '+%Y-%m-%d %H:%M:%S')] ✅ 已推送无风险状态卡片" >> "$LOG_FILE"
+        exit 0
+    fi
+fi
+
+# Step 2: 生成 AI 任务文件
+echo "[$(date '+%Y-%m-%d %H:%M:%S')] 步骤2: 生成 AI 分析任务..." >> "$LOG_FILE"
+$NODE_PATH risk-alert.js >> "$LOG_FILE" 2>&1
+
+# Step 3: 调用 AI 分析（通过发送消息触发）
+echo "[$(date '+%Y-%m-%d %H:%M:%S')] 步骤3: 等待 AI 分析..." >> "$LOG_FILE"
+echo "请 AI 助手执行风险分析并生成 risk-data.json，然后执行: node risk-alert.js push"
+
+# 注意：AI 分析需要手动触发或由 AI 自动处理
+# 这里我们直接推送（假设 AI 已经完成分析或 risk-data.json 已存在）
+
+# Step 4: 推送卡片
+echo "[$(date '+%Y-%m-%d %H:%M:%S')] 步骤4: 推送卡片..." >> "$LOG_FILE"
+$NODE_PATH risk-alert.js push >> "$LOG_FILE" 2>&1
+if [ $? -eq 0 ]; then
+    echo "[$(date '+%Y-%m-%d %H:%M:%S')] ✅ 推送成功" >> "$LOG_FILE"
+else
+    echo "[$(date '+%Y-%m-%d %H:%M:%S')] ❌ 推送失败" >> "$LOG_FILE"
+fi
+
+echo "[$(date '+%Y-%m-%d %H:%M:%S')] 定时任务执行完毕" >> "$LOG_FILE"
+echo "---" >> "$LOG_FILE"