Skip to content

ZJU-REAL/Easel / skills/builtin/skill-wechat-publisher

skill-wechat-publisher

Verified

微信公众号文章自动创作与发布工具。给定参考文章、文字或文档,自动搜索整理全网相关信息、生成图文并茂的公众号文章,并发布到微信公众号草稿箱。特别强调反 AI 检测写作。 触发场景(沾边就用):用户提到"公众号 / 微信文章 / 推文 / 公号 / 发文 / mp / 草稿箱 / 群发",或要求把话题/文档/笔记写成公众号文章并发布到微信。

FileKindSize
EASEL-META.mdmarkdown2.6 KB
SKILL.mdskill md11.4 KB
assets/image-styles/README.mdmarkdown29.5 KB
assets/image-styles/blue.jsonconfig3.0 KB
assets/image-styles/data-chart.jsonconfig2.4 KB
assets/image-styles/flat-editorial.jsonconfig4.9 KB
assets/image-styles/hand-drawn-blue.jsonconfig2.3 KB
assets/image-styles/illustrated-warm.jsonconfig3.3 KB
assets/image-styles/infographic-blue.jsonconfig5.2 KB
assets/image-styles/infographic-dark.jsonconfig5.4 KB
assets/image-styles/infographic-mint.jsonconfig5.2 KB
assets/image-styles/infographic-warm.jsonconfig5.4 KB
assets/image-styles/knowledge-card.jsonconfig2.3 KB
assets/image-styles/magazine-editorial.jsonconfig2.3 KB
assets/image-styles/marker-coral.jsonconfig6.4 KB
assets/image-styles/marker-lime.jsonconfig6.6 KB
assets/image-styles/marker-pink.jsonconfig6.4 KB
assets/image-styles/marker-sky.jsonconfig6.4 KB
assets/image-styles/marker-violet.jsonconfig6.4 KB
assets/image-styles/meme-illustration.jsonconfig2.3 KB
assets/image-styles/mint.jsonconfig3.1 KB
assets/image-styles/morandi.jsonconfig3.4 KB
assets/image-styles/previews/blue.webpbinary104.0 KB
assets/image-styles/previews/data-chart.webpbinary16.4 KB
assets/image-styles/previews/flat-editorial.webpbinary109.7 KB
assets/image-styles/previews/hand-drawn-blue.webpbinary17.9 KB
assets/image-styles/previews/illustrated-warm.webpbinary31.2 KB
assets/image-styles/previews/infographic-blue.webpbinary160.4 KB
assets/image-styles/previews/infographic-dark.webpbinary142.1 KB
assets/image-styles/previews/infographic-mint.webpbinary150.2 KB
assets/image-styles/previews/infographic-warm.webpbinary162.5 KB
assets/image-styles/previews/knowledge-card.webpbinary22.4 KB
assets/image-styles/previews/magazine-editorial.webpbinary14.0 KB
assets/image-styles/previews/marker-coral.webpbinary41.2 KB
assets/image-styles/previews/marker-lime-template.webpbinary34.3 KB
assets/image-styles/previews/marker-lime.webpbinary39.3 KB
assets/image-styles/previews/marker-pink.webpbinary37.5 KB
assets/image-styles/previews/marker-sky.webpbinary47.9 KB
assets/image-styles/previews/marker-violet.webpbinary42.8 KB
assets/image-styles/previews/meme-illustration.webpbinary35.5 KB
assets/image-styles/previews/mint.webpbinary95.7 KB
assets/image-styles/previews/morandi.webpbinary82.0 KB
assets/image-styles/previews/slice-of-life.webpbinary46.3 KB
assets/image-styles/previews/warm-handdrawn.webpbinary79.0 KB
assets/image-styles/previews/xiaohongshu-colorful.webpbinary24.3 KB
assets/image-styles/slice-of-life.jsonconfig4.9 KB
assets/image-styles/warm-handdrawn.jsonconfig4.4 KB
assets/image-styles/xiaohongshu-colorful.jsonconfig2.3 KB
assets/theme-previews/README.mdmarkdown1.7 KB
assets/theme-previews/academic-paper.htmlother13.5 KB
assets/theme-previews/business-navy.htmlother13.6 KB
assets/theme-previews/elegant-ink.htmlother13.0 KB
assets/theme-previews/girly-pink.htmlother13.2 KB
assets/theme-previews/index.htmlother26.6 KB
assets/theme-previews/ink-wash.htmlother13.5 KB
assets/theme-previews/magazine-grid.htmlother13.0 KB
assets/theme-previews/minimal-bw.htmlother13.6 KB
assets/theme-previews/minimal-mono.htmlother12.7 KB
assets/theme-previews/mint-fresh.htmlother14.0 KB
assets/theme-previews/news-bold.htmlother13.3 KB
assets/theme-previews/refined-blue.htmlother15.1 KB
assets/theme-previews/sage-premium.htmlother13.6 KB
assets/theme-previews/sample.mdmarkdown2.6 KB
assets/theme-previews/screenshots/ink-wash.webpbinary27.8 KB
assets/theme-previews/screenshots/minimal-mono.webpbinary34.8 KB
assets/theme-previews/screenshots/news-bold.webpbinary32.5 KB
assets/theme-previews/screenshots/refined-blue.webpbinary34.3 KB
assets/theme-previews/screenshots/theme-overview.webpbinary229.8 KB
assets/theme-previews/sunset-coral.htmlother14.8 KB
assets/theme-previews/warm-editorial.htmlother12.8 KB
assets/theme-previews/warm-orange.htmlother12.3 KB
assets/themes/academic-paper.jsonconfig5.2 KB
assets/themes/business-navy.jsonconfig5.2 KB
assets/themes/elegant-ink.jsonconfig5.1 KB
assets/themes/girly-pink.jsonconfig5.3 KB
assets/themes/ink-wash.jsonconfig5.1 KB
assets/themes/magazine-grid.jsonconfig5.1 KB
assets/themes/minimal-bw.jsonconfig5.2 KB
assets/themes/minimal-mono.jsonconfig5.0 KB
assets/themes/mint-fresh.jsonconfig5.3 KB
assets/themes/news-bold.jsonconfig5.1 KB
assets/themes/refined-blue.jsonconfig5.6 KB
assets/themes/sage-premium.jsonconfig5.3 KB
assets/themes/sunset-coral.jsonconfig5.5 KB
assets/themes/warm-editorial.jsonconfig4.9 KB
assets/themes/warm-orange.jsonconfig5.0 KB
assets/typeset-card/template.htmlother3.5 KB
brief.md.exampleother2.3 KB
references/anti-ai-checklist.mdmarkdown3.4 KB
references/api_reference.mdmarkdown3.2 KB
references/article-structures.mdmarkdown3.2 KB
references/errors.mdmarkdown1.1 KB
references/image-styles-guide.mdmarkdown3.5 KB
references/inline-markup.mdmarkdown1.7 KB
references/multi-platform-sync.mdmarkdown2.6 KB
references/newspic-mode.mdmarkdown5.7 KB
references/themes.mdmarkdown1.8 KB
references/typeset-card.mdmarkdown4.3 KB
scripts/ai_score.pyexecutablescript15.2 KB
scripts/api.pyexecutablescript13.3 KB
scripts/baoyu_image_gen.tsexecutablescript5.5 KB
scripts/baoyu_image_gen_core.tsscript11.5 KB
scripts/config.pyexecutablescript11.7 KB
scripts/generate_image.pyexecutablescript4.3 KB
scripts/html_converter.pyexecutablescript23.3 KB
scripts/image_handler.pyexecutablescript17.5 KB
scripts/multi_publish.pyexecutablescript14.2 KB
scripts/newspic_build.pyexecutablescript14.0 KB
scripts/publish.pyexecutablescript35.7 KB
scripts/sync_pages.shexecutablescript994 B
scripts/wechat_api.pyexecutablescript6.0 KB
scripts/wechat_token.pyexecutablescript5.1 KB
tests/__init__.pyscript0 B
tests/conftest.pyscript4.1 KB
tests/test_ai_score.pyscript8.1 KB
tests/test_config.pyscript7.4 KB
tests/test_config_integration.pyscript9.7 KB
tests/test_draft_submission.pyscript3.1 KB
tests/test_html_converter.pyscript10.2 KB
wechat-publisher.yaml.exampleother1.8 KB

scripts/publish.py

View on GitHub
#!/usr/bin/env python3
"""
微信公众号文章发布 - 主流程脚本

整合所有模块,实现从Markdown文章到微信公众号草稿箱的一键发布:
1. 读取Markdown文章
2. 处理文章中的图片(下载+上传到微信)
3. 转换为微信兼容HTML(内联样式排版)
4. 准备封面图
5. 调用API创建草稿

用法:
    python publish.py --input article.md --author "作者名"
    python publish.py --input article.md --cover cover.jpg --title "自定义标题"
    python publish.py --html article.html --cover cover.jpg --title "标题"
"""

import sys
import re
import json
import argparse
import tempfile
import uuid
from pathlib import Path
from typing import Optional

# 添加脚本目录到路径
sys.path.insert(0, str(Path(__file__).parent))
sys.path.insert(0, str(Path(__file__).resolve().parents[3] / "shared" / "scripts"))

import content_guard

from wechat_api import (
    publish_article,
    publish_newspic,
    upload_thumb_image,
    get_access_token,
    set_account,
    get_config,
    ConfigError,
    resolve_image_style,
    list_image_styles,
)
from html_converter import convert_markdown_to_wechat_html, load_styles, load_theme
from image_handler import process_article_images, extract_images_from_markdown


# ============================================================
# Markdown 文本处理工具
# ============================================================

# 成对行内标记(标记符必须成对出现),共用一个替换函数
_INLINE_PAIR_MARKERS = [
    (r"\*\*", r"\*\*"),   # **粗体**
    (r"==",   r"=="),      # ==黄色高亮==
    (r"\+\+", r"\+\+"),   # ++蓝色高亮++
    (r"%%",   r"%%"),      # %%粉色高亮%%
    (r"&&",   r"&&"),      # &&绿色高亮&&
    (r"!!",   r"!!"),      # !!红色强调!!
    (r"@@",   r"@@"),      # @@蓝色强调@@
    (r"\^\^", r"\^\^"),   # ^^橙色强调^^
]


def _strip_inline_markers(text: str) -> str:
    """剥离 Markdown 行内排版标记(加粗/斜体/自定义标色),保留文字内容本身。

    处理成对标记(比如 `**foo**` → `foo`),同时清除末尾没配对的孤立符号,
    避免摘要里冒出 `**` 之类的裸标记。
    """
    # 先处理成对标记
    for open_pat, close_pat in _INLINE_PAIR_MARKERS:
        text = re.sub(rf"{open_pat}([^\n]+?){close_pat}", r"\1", text)
    # 斜体(单星号)
    text = re.sub(r"\*([^*\n]+)\*", r"\1", text)
    # 剩余的孤立标记符号一律剥掉,避免摘要里冒出 `**` 之类
    text = re.sub(r"\*\*|==|\+\+|%%|&&|!!|@@|\^\^", "", text)
    # 反引号
    text = text.replace("`", "")
    return text


def _strip_front_matter(md_content: str) -> str:
    """如果文章以 YAML front matter(--- ... ---)开头,去掉整段 front matter。"""
    if md_content.lstrip().startswith("---"):
        # 找到起始 `---` 后的第二个 `---` 行
        m = re.match(r"^\s*---\s*\n(.*?)\n---\s*\n", md_content, flags=re.DOTALL)
        if m:
            return md_content[m.end():]
    return md_content


def extract_title_from_markdown(md_content):
    """从Markdown中提取标题(第一个#标题)。

    回退顺序:
      1. 第一个 `# ` 标题
      2. 跳过 YAML front matter 后的第一行非空文本(排除引用/列表/图片等)
    """
    match = re.search(r'^#\s+(.+)$', md_content, re.MULTILINE)
    if match:
        return _strip_inline_markers(match.group(1).strip())

    # 回退前先剥掉 YAML front matter,避免把 `---` 或 `title: xxx` 当成标题
    body = _strip_front_matter(md_content)
    for line in body.split("\n"):
        line = line.strip()
        if not line:
            continue
        # 排除引用、图片、分隔线、列表项
        if line.startswith((">", "!", "-", "*", "+")):
            continue
        if set(line) <= {"-", "=", " "}:  # 纯分隔线
            continue
        return _strip_inline_markers(line)[:50]
    return "未命名文章"


def extract_digest_from_markdown(md_content):
    """从Markdown中提取摘要(blockquote或前120字)。

    清理所有行内排版标记(`**foo**`, `==foo==` 等)后再截断,避免摘要里
    残留 `**`、`==` 这类裸标记符号。
    """
    # 优先使用 blockquote
    match = re.search(r'^>\s+(.+)$', md_content, re.MULTILINE)
    if match:
        cleaned = _strip_inline_markers(match.group(1).strip())
        return cleaned[:120]

    # 回退:提取正文前120字
    text = _strip_front_matter(md_content)
    # 去掉标题符号、引用符号、图片链接
    text = re.sub(r'^#+\s*', '', text, flags=re.MULTILINE)
    text = re.sub(r'^>\s*', '', text, flags=re.MULTILINE)
    text = re.sub(r'!\[[^\]]*\]\([^)]+\)', '', text)
    text = re.sub(r'\[([^\]]+)\]\([^)]+\)', r'\1', text)
    text = _strip_inline_markers(text)
    # 剥掉剩余的方括号
    text = re.sub(r'[\[\]]', '', text)
    text = " ".join(text.split())
    return text[:120]


def remove_title_from_content(md_content):
    """从内容中移除标题行(避免在正文中重复显示标题)"""
    lines = md_content.split("\n")
    result = []
    title_removed = False
    for line in lines:
        if not title_removed and re.match(r'^#\s+', line.strip()):
            title_removed = True
            continue
        result.append(line)
    return "\n".join(result)


# ============================================================
# 默认临时目录
# ============================================================

def _default_temp_dir() -> str:
    """为本次运行创建一个独立的临时目录,避免并发运行互相覆盖文件。"""
    return tempfile.mkdtemp(prefix="wechat_images_")


def _session_publish_html(html_content, cover_path, title, digest="", author="", base_dir=None):
    """会话式发布:把 HTML 写进 base_dir(让正文相对图能被解析),调
    shared/scripts/weixin_mp_stats.py publish(走公众号后台会话,免 app_secret/免 IP 白名单,
    正文内嵌图自动传 mp CDN)。返回 {media_id,...}。"""
    import subprocess as _sp
    import os as _os
    shared = Path(__file__).resolve().parents[3] / "shared" / "scripts"
    base = Path(base_dir) if base_dir else Path(_default_temp_dir())
    base.mkdir(parents=True, exist_ok=True)
    # 只需一个随机文件名,别再调 _default_temp_dir()(它会 mkdtemp 真建目录,
    # 每次发布遗留一个空的 wechat_images_XXXX/)。用 uuid 取随机名即可。
    html_tmp = base / f".session_{uuid.uuid4().hex}.html"
    html_tmp.write_text(html_content, encoding="utf-8")
    # 默认直连(mp 为国内站);受限网络可用 EASEL_PROXY / https_proxy 指定正向代理
    proxy = _os.environ.get("EASEL_PROXY") or _os.environ.get("https_proxy") or ""
    cmd = [sys.executable, str(shared / "weixin_mp_stats.py"), "publish", "--proxy", proxy,
           "--html", str(html_tmp), "--cover", str(cover_path), "--title", title,
           "--digest", digest or "", "--author", author or ""]
    proc = _sp.run(cmd, capture_output=True, text=True, timeout=300)
    for line in reversed((proc.stdout or "").strip().splitlines()):
        line = line.strip()
        if line.startswith("{"):
            try:
                res = json.loads(line)
            except Exception:
                continue
            if res.get("success"):
                return {"media_id": res.get("media_id", ""), "status": "success",
                        "via": "session", "thumb_media_id": res.get("thumb_media_id", "")}
            raise RuntimeError(f"会话式发布失败:{res.get('error') or res}")
    raise RuntimeError(f"会话式发布无有效返回:{(proc.stderr or proc.stdout or '')[-300:]}")


# ============================================================
# 发布主流程
# ============================================================

def publish_from_markdown(
    md_path,
    title=None,
    author=None,
    digest=None,
    cover_path=None,
    source_url="",
    temp_dir=None,
    style_path=None,
    theme=None,
    sync_platforms=None,
    account_name: Optional[str] = None,
    ai_score_threshold: float = None,  # None → use ai_score.DEFAULT_THRESHOLD
    skip_ai_score: bool = False,
    allow_missing_images: bool = False,
    session_mode: bool = True,
    debug: bool = False,
):
    """
    从Markdown文件发布文章到微信公众号草稿箱。

    完整流程:
    1. 读取Markdown
    2. AI 味检测(gate,可用 skip_ai_score 关闭)
    3. 提取/确认标题和摘要
    4. 处理图片(下载+上传微信)
    5. 转换HTML排版
    6. 准备封面图
    7. 创建草稿

    Args:
        md_path: Markdown文件路径
        title: 标题(可选,默认从Markdown提取)
        author: 作者名(默认从账号配置读取,再兜底到 "飞哥")
        digest: 摘要(可选,默认从Markdown提取)
        cover_path: 封面图路径(可选,默认使用文章第一张图)
        source_url: 原文链接
        temp_dir: 临时文件目录(默认为每次运行独立生成)
        style_path: 自定义样式文件路径
        theme: 排版主题名(默认从账号配置读取)
        sync_platforms: 同步到其他平台的列表
        account_name: 若指定,则在函数内部调用 set_account() 切换账号
        ai_score_threshold: AI 味检测阈值,总分 >= 阈值视为不通过(默认 45)
        skip_ai_score: True 时跳过 AI 味检测(默认 False)
        allow_missing_images: True 时允许部分图片上传失败仍继续发布(默认 False,缺图即中止)
        debug: True 时把生成的 HTML 保存到 temp_dir 下方便调试

    Returns:
        dict: 发布结果
    """
    # 如果库调用方指定了账号,立即切换
    if account_name:
        set_account(account_name)

    if temp_dir is None:
        temp_dir = _default_temp_dir()

    print("=" * 60)
    print("微信公众号文章发布")
    print("=" * 60)

    # 1. 读取文章
    md_path = Path(md_path)
    if not md_path.exists():
        raise FileNotFoundError(f"文章文件不存在:{md_path}")

    with open(md_path, "r", encoding="utf-8") as f:
        md_content = f.read()

    print(f"读取文章:{md_path} ({len(md_content)} 字符)")

    # 2. AI 味检测(在图片上传前,失败时避免浪费素材库配额)
    if not skip_ai_score:
        # 延迟导入,避免其他 agent 正在写 ai_score.py 时触发 import 错误
        try:
            from ai_score import check_ai_score, DEFAULT_THRESHOLD as _DEFAULT_THRESHOLD
        except ImportError as exc:
            print(f"\n[警告] 无法加载 ai_score 模块({exc}),跳过 AI 味检测。")
        else:
            # 使用 ai_score.DEFAULT_THRESHOLD 作为默认值,保证 CLI 与库行为一致
            _threshold = ai_score_threshold if ai_score_threshold is not None else _DEFAULT_THRESHOLD
            print("\n[预检] AI 味检测...")
            passed, report = check_ai_score(md_content, _threshold)
            total = report.get("total_score")
            print(f"  总分: {total} / 100  (阈值 {_threshold})")
            if not passed:
                print(f"  结果: 未通过 AI 味检测(总分 {total} >= 阈值 {_threshold})")
                dims = report.get("dimensions", {}) or {}
                if dims:
                    print("  失分维度:")
                    for name, info in dims.items():
                        score = info.get("score") if isinstance(info, dict) else info
                        print(f"    · {name}: {score}")
                hit_phrases = report.get("hit_phrases") or []
                if hit_phrases:
                    print("  命中 AI 套话:")
                    for s in hit_phrases[:10]:
                        print(f"    · {s}")
                hit_vocab = report.get("hit_vocab") or []
                if hit_vocab:
                    print("  命中 AI 高频词: " + ", ".join(str(v) for v in hit_vocab[:15]))
                print(
                    "\n  发布已阻止。请回到 SKILL.md 阶段 3.5『人味化改写』按清单逐条打磨,\n"
                    "  或使用 --skip-ai-score 跳过此检测(不推荐)。"
                )
                raise SystemExit(1)
            print("  结果: 通过")

    # 3. 提取标题和摘要
    if not title:
        title = extract_title_from_markdown(md_content)
    if not digest:
        digest = extract_digest_from_markdown(md_content)

    print(f"标题:{title}")
    print(f"摘要:{digest[:50]}...")

    content_guard.guard_or_die(
        [title, digest, md_content, source_url],
        exec_mode=True,
        label="微信公众号草稿内容",
    )

    # 3.5 会话式发布(默认):MD→公众号HTML(保留本地图引用) → 后台会话建草稿,免 app_secret / 免 IP 白名单。
    #      正文内嵌图与封面由会话发布器自行上传 mp CDN;不走官方 API 的 process_article_images / draft_add。
    if session_mode:
        print("\n[会话式发布] 走公众号后台会话(免 AppID/IP 白名单)...")
        styles, highlights, divider_text, list_style = load_theme(theme_name=theme, style_path=style_path)
        content_md = remove_title_from_content(md_content)
        html_content = convert_markdown_to_wechat_html(content_md, styles, highlights, divider_text, list_style)
        cov = cover_path
        if not cov:
            imgs = extract_images_from_markdown(content_md)
            cov = imgs[0]["url"] if imgs else None
            if cov and not Path(cov).is_absolute():
                cov = str((md_path.parent / cov))
        if not cov or not Path(cov).exists():
            print("  错误:需要封面图(--cover 或正文第一张本地图片)"); sys.exit(1)
        result = _session_publish_html(html_content, cov, title, digest, author or "",
                                       base_dir=md_path.parent)
        print("\n" + "=" * 60)
        print("发布完成(会话式)!")
        print(f"  草稿 media_id: {result['media_id']}")
        print("  请登录微信公众平台查看草稿箱")
        print("=" * 60)
        return result

    # 4. 验证API连接
    print("\n[步骤1] 验证API连接...")
    try:
        token = get_access_token()
        print(f"  API连接正常,token已获取")
    except Exception as e:
        print(f"  API连接失败:{e}")
        print("  请检查 wechat-publisher.yaml 中的 app_id / app_secret 配置")
        sys.exit(1)

    # 5. 处理图片
    print("\n[步骤2] 处理文章图片...")
    # 移除标题后处理(标题不包含在正文中)
    content_md = remove_title_from_content(md_content)
    processed_md, img_mapping, first_img = process_article_images(
        content_md, temp_dir, base_dir=md_path.parent
    )

    # 缺图守卫:引用了但未成功上传的图片,默认中止发布,避免产出缺图草稿
    referenced = {img["url"] for img in extract_images_from_markdown(content_md)}
    missing = sorted(u for u in referenced if u not in img_mapping)
    if missing:
        print(f"\n  错误:{len(missing)} 张图片引用未能上传:")
        for u in missing:
            print(f"    · {u}")
        if allow_missing_images:
            print("  --allow-missing-images 已指定,继续发布(草稿将缺图)。")
        else:
            print(
                "  发布已中止。请检查图片路径(相对路径按 article.md 所在目录解析),\n"
                "  或显式传 --allow-missing-images 接受缺图草稿。"
            )
            raise SystemExit(1)

    # 6. 转换HTML
    print("\n[步骤3] 转换HTML排版...")
    styles, highlights, divider_text, list_style = load_theme(
        theme_name=theme, style_path=style_path
    )
    print(f"  主题: {theme or '(默认)'}")
    html_content = convert_markdown_to_wechat_html(
        processed_md, styles, highlights, divider_text, list_style
    )
    print(f"  HTML生成完成 ({len(html_content)} 字符)")

    # 仅在 debug 模式下持久化中间 HTML(否则会堆满临时目录)
    if debug:
        html_path = Path(temp_dir) / "article_output.html"
        html_path.parent.mkdir(parents=True, exist_ok=True)
        with open(html_path, "w", encoding="utf-8") as f:
            f.write(html_content)
        print(f"  [debug] HTML已保存到:{html_path}")

    # 7. 准备封面图
    print("\n[步骤4] 准备封面图...")
    if cover_path and Path(cover_path).exists():
        print(f"  使用指定封面图:{cover_path}")
    elif first_img:
        cover_path = first_img
        print(f"  使用文章第一张图作为封面:{cover_path}")
    else:
        print("  警告:没有封面图!请提供 --cover 参数")
        print("  微信公众号要求每篇文章必须有封面图")
        sys.exit(1)

    # 8. 如未提供 author,从账号配置兜底读取
    if author is None:
        try:
            cfg = get_config(account_name)
            author = cfg.get("author", "") or "飞哥"
        except ConfigError:
            author = "飞哥"

    # 9. 发布到草稿箱
    print("\n[步骤5] 发布到草稿箱...")
    result = publish_article(
        title=title,
        html_content=html_content,
        cover_image_path=cover_path,
        author=author,
        digest=digest,
        source_url=source_url,
    )

    print("\n" + "=" * 60)
    print("发布完成!")
    print(f"  草稿 media_id: {result['media_id']}")
    print("  请登录微信公众平台查看草稿箱")
    print("=" * 60)

    # 可选阶段:同步到其他内容平台(opt-in,失败不影响微信发布结果)
    if sync_platforms:
        print(f"\n[可选] 同步到其他平台: {', '.join(sync_platforms)}")
        try:
            from multi_publish import run as sync_run
            sync_result = sync_run(
                md_path=md_path,
                platforms=sync_platforms,
                title=title,
                cover_path=cover_path,
                strict=False,
            )
            result["sync"] = {
                "platforms": sync_platforms,
                "success": sync_result["success"],
                "returncode": sync_result["returncode"],
            }
        except Exception as e:
            print(f"  同步环节异常(不影响微信发布): {e}")
            result["sync"] = {"platforms": sync_platforms, "success": False, "error": str(e)}

    return result


def publish_from_brief(
    brief_path,
    account_name: Optional[str] = None,
    image_style: Optional[str] = None,
    title: Optional[str] = None,
    author: Optional[str] = None,
    ai_score_threshold: float = None,
    skip_ai_score: bool = False,
):
    """
    从 brief.md 发布贴图(newspic)到草稿箱。

    期望目录布局(brief.md 旁边):
        <dir>/
          brief.md        # 本次输入,含 frontmatter + 要点 + 短文本
          card_plan.json  # (可选)newspic_build.py 生成过就存在
          images/
            01.png
            02.png
            ...

    完整流程:
      1. 解析 brief.md,拿到 topic / title / account / image_style / 短文本
      2. 检查 images/ 下按 NN.png(或其他格式)顺序命名的图片
      3. 如短文本不空,过 AI 味 gate(newspic 模式)
      4. 依次上传图片为永久素材,收到 media_id
      5. 调 draft/add article_type=newspic 建草稿

    Args:
        brief_path: brief.md 路径
        account_name: 覆盖 frontmatter.account
        image_style: 覆盖 frontmatter.image_style(优先级最高,但本函数主要用于发布,
                     真正影响生图的是 newspic_build.py。这里仅记录到返回值里)
        title: 覆盖 frontmatter.title
        author: 覆盖账号默认作者
        ai_score_threshold: AI 味检测阈值(newspic 短文本)
        skip_ai_score: 跳过短文本检测

    Returns:
        dict: 发布结果(同 publish_newspic 返回值),外加 "image_style" / "cards" 信息
    """
    # 延迟导入以避免与 newspic_build 的循环依赖(实际无依赖,但保持对称)
    from newspic_build import parse_brief

    brief_path = Path(brief_path)
    if not brief_path.exists():
        raise FileNotFoundError(f"brief.md 不存在: {brief_path}")

    parsed = parse_brief(brief_path.read_text(encoding="utf-8"))
    fm = parsed["frontmatter"] or {}
    sections = parsed["sections"]

    # 账号解析: CLI > frontmatter > 全局激活
    final_account = account_name or fm.get("account")
    if final_account:
        set_account(final_account)

    # 标题/作者/风格(贴图模式走 newspic,让兜底用手绘水彩信息图而非文章线条手绘)
    final_title = title or fm.get("title", "") or ""
    style = resolve_image_style(
        cli_value=image_style,
        frontmatter_value=fm.get("image_style"),
        account_name=final_account,
        mode="newspic",
    )

    # 短文本
    short_text = (sections.get("短文本") or sections.get("短描述") or "").strip()

    print("=" * 60)
    print("微信公众号贴图(newspic)发布")
    print("=" * 60)
    print(f"brief:    {brief_path}")
    print(f"话题:     {fm.get('topic', '(未填)')}")
    print(f"标题:     {final_title or '(空)'}")
    print(f"风格:     {style['style_name']} ({style.get('display_name','')})")
    print(f"短文本:   {len(short_text)} 字")

    content_guard.guard_or_die(
        [final_title, short_text],
        exec_mode=True,
        label="微信公众号贴图草稿内容",
    )

    # 收集图片
    images_dir = brief_path.parent / "images"
    if not images_dir.exists():
        raise FileNotFoundError(
            f"找不到图片目录: {images_dir}\n"
            f"请先用 newspic_build.py + scripts/baoyu_image_gen.ts 生成贴图,保存为 {images_dir}/01.png 起。"
        )
    exts = (".png", ".jpg", ".jpeg", ".webp")
    image_paths = sorted(
        [p for p in images_dir.iterdir() if p.suffix.lower() in exts],
        key=lambda p: p.name,
    )
    if not image_paths:
        raise FileNotFoundError(f"{images_dir} 下没有图片,请先生图")
    if len(image_paths) < 2:
        print(f"[警告] 只有 {len(image_paths)} 张图。newspic 建议 5-10 张。")
    if len(image_paths) > 20:
        print(f"[警告] {len(image_paths)} 张超出微信 20 张上限,截取前 20 张")
        image_paths = image_paths[:20]
    print(f"图片:     {len(image_paths)} 张  (首张 {image_paths[0].name} 为封面)")
    print("-" * 60)

    # AI 味 gate(newspic 模式)
    if not skip_ai_score and short_text:
        try:
            from ai_score import check_ai_score, DEFAULT_THRESHOLD as _DEFAULT_THRESHOLD
        except ImportError as exc:
            print(f"[警告] 无法加载 ai_score({exc}),跳过短文本检测。")
        else:
            _threshold = ai_score_threshold if ai_score_threshold is not None else _DEFAULT_THRESHOLD
            print(f"\n[预检] AI 味检测(newspic 模式,阈值 {_threshold})...")
            passed, report = check_ai_score(short_text, _threshold, mode="newspic")
            print(f"  总分: {report['total_score']} / 100")
            if not passed:
                print(f"  结果: 未通过(>= 阈值 {_threshold})")
                hit_phrases = report.get("hit_phrases") or []
                hit_vocab = report.get("hit_vocab") or []
                if hit_phrases:
                    print("  命中 AI 套话: " + ", ".join(hit_phrases[:8]))
                if hit_vocab:
                    print("  命中 AI 高频词: " + ", ".join(str(v) for v in hit_vocab[:10]))
                print(
                    "\n  发布已阻止。请重写 brief.md 的『# 短文本』段落,去掉套话。\n"
                    "  或用 --skip-ai-score 跳过(不推荐)。"
                )
                raise SystemExit(1)
            print("  结果: 通过")

    # 验证 API 连接(fail fast,避免图片上传一半才发现 token 错)
    print("\n[步骤1] 验证 API 连接...")
    try:
        get_access_token()
        print("  API 连接正常")
    except Exception as e:
        print(f"  API 连接失败: {e}")
        raise SystemExit(1)

    # 作者兜底
    if author is None:
        try:
            cfg = get_config(final_account)
            author = cfg.get("author", "") or ""
        except ConfigError:
            author = ""

    # 上传 + 建草稿
    print("\n[步骤2] 上传图片 + 创建贴图草稿...")
    result = publish_newspic(
        title=final_title,
        content=short_text,
        image_paths=image_paths,
        author=author,
    )

    result["image_style"] = style["style_name"]
    result["type"] = "newspic"
    result["cards"] = len(image_paths)
    result["brief_path"] = str(brief_path)

    print("\n" + "=" * 60)
    print("贴图发布完成!")
    print(f"  草稿 media_id: {result['media_id']}")
    print(f"  图片张数: {result['cards']}")
    print(f"  风格: {result['image_style']}")
    print("  请登录微信公众平台查看草稿箱")
    print("=" * 60)

    return result


def publish_from_html(
    html_path,
    title,
    cover_path,
    author=None,
    digest="",
    source_url="",
    account_name: Optional[str] = None,
):
    """
    从已排版的HTML文件发布文章。

    适用于已经完成排版的HTML内容,直接上传。HTML 已经是微信专用排版,
    因此不做 AI 味检测,也不支持多平台同步。

    Args:
        html_path: HTML文件路径
        title: 文章标题(必需)
        cover_path: 封面图路径(必需)
        author: 作者名(默认从账号配置读取,再兜底到 "飞哥")
        digest: 摘要
        source_url: 原文链接
        account_name: 若指定,则在函数内部调用 set_account() 切换账号

    Returns:
        dict: 发布结果
    """
    if account_name:
        set_account(account_name)

    with open(html_path, "r", encoding="utf-8") as f:
        html_content = f.read()

    content_guard.guard_or_die(
        [title, digest, html_content, source_url],
        exec_mode=True,
        label="微信公众号 HTML 草稿内容",
    )

    # 如未提供 author,从账号配置兜底读取
    if author is None:
        try:
            cfg = get_config(account_name)
            author = cfg.get("author", "") or "飞哥"
        except ConfigError:
            author = "飞哥"

    result = publish_article(
        title=title,
        html_content=html_content,
        cover_image_path=cover_path,
        author=author,
        digest=digest,
        source_url=source_url,
    )
    return result


# ============================================================
# 命令行入口
# ============================================================

def _build_parser() -> argparse.ArgumentParser:
    parser = argparse.ArgumentParser(
        description="微信公众号文章一键发布到草稿箱",
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog="""
示例:
  # 从Markdown发布
  python publish.py --input article.md --author "张三"

  # 指定封面图和标题
  python publish.py --input article.md --cover cover.jpg --title "自定义标题"

  # 从HTML发布
  python publish.py --html article.html --cover cover.jpg --title "标题"

  # 跳过 AI 味检测或放宽阈值
  python publish.py --input article.md --cover cover.jpg --skip-ai-score
  python publish.py --input article.md --cover cover.jpg --ai-score-threshold 55
        """
    )

    parser.add_argument("--input", "-i", help="Markdown文件路径")
    parser.add_argument("--html", help="HTML文件路径(已排版)")
    parser.add_argument("--brief", help="贴图(newspic)模式的 brief.md 路径")
    parser.add_argument(
        "--type",
        choices=["news", "newspic"],
        default=None,
        help="发布类型: news(图文,默认) / newspic(贴图/图片消息)。"
             "--brief 存在时自动推断为 newspic",
    )
    parser.add_argument("--title", "-t", help="文章标题(默认从文章提取)")
    parser.add_argument("--cover", "-c", help="封面图路径")
    parser.add_argument("--author", "-a", default=None,
                        help="作者名(默认从账号配置获取,兜底:飞哥)")
    parser.add_argument("--digest", "-d", help="文章摘要(默认从文章提取)")
    parser.add_argument("--source-url", default="", help="原文链接")
    parser.add_argument("--style", help="自定义样式JSON路径")
    parser.add_argument("--theme", help="排版主题名(对应 assets/themes/<name>.json,默认从账号配置读取)")
    parser.add_argument(
        "--image-style",
        help="配图风格名(对应 assets/image-styles/<name>.json)。"
             "优先级: CLI > brief/article frontmatter > 账号默认 > hand-drawn-blue。"
             "list: python3 wechat_api.py list-image-styles",
    )
    parser.add_argument(
        "--temp-dir",
        default=None,
        help="临时文件目录(默认每次运行独立生成,避免并发冲突)",
    )
    parser.add_argument("--account", help="指定公众号账号(对应 wechat-publisher.yaml 中的账号名)")
    parser.add_argument(
        "--sync",
        help="可选:发到微信后同步到其他平台,逗号分隔(如 zhihu,juejin,csdn)。"
             "覆盖 --sync-from-config。需先装 @wechatsync/cli 并配置 WECHATSYNC_MCP_TOKEN",
    )
    parser.add_argument(
        "--sync-from-config", action="store_true",
        help="可选:同步到账号配置里 sync_platforms 字段指定的平台",
    )
    parser.add_argument(
        "--ai-score-threshold",
        type=float,
        default=None,
        help="AI 味检测阈值(默认与 ai_score.DEFAULT_THRESHOLD 一致)。总分 >= 阈值则阻止发布",
    )
    parser.add_argument(
        "--skip-ai-score",
        action="store_true",
        help="跳过 AI 味检测(不推荐,仅在已经人工改写过时使用)",
    )
    parser.add_argument(
        "--allow-missing-images",
        action="store_true",
        help="允许部分图片上传失败仍继续发布(默认缺图即中止,避免产出缺图草稿)",
    )
    parser.add_argument(
        "--debug",
        action="store_true",
        help="调试模式:把生成的 HTML 保存到 temp_dir 方便排查",
    )
    parser.add_argument(
        "--exec",
        action="store_true",
        help="确认创建微信草稿;默认仅预览将使用的账号、模式和输入文件",
    )
    parser.add_argument(
        "--official-api",
        action="store_true",
        help="回退到官方 API 发布(需 app_id/app_secret + IP 白名单);"
             "默认走公众号后台会话发布(免 secret、免白名单,需先扫码登录后台)",
    )
    return parser


def _resolve_config(args):
    """按命令行参数回读账号配置,填充 author/theme/sync_platforms。

    返回 config_sync_platforms(配置里声明的同步平台列表,可能为 None)。
    约定:
      - `--sync-from-config` 显式要求读配置,任何 ConfigError 都应 exit 1。
      - 其他情况(只是兜底 author/theme)下 ConfigError 会被静默吞掉,
        保留原有"没有配置文件也能用 --author 硬写"的行为。
    """
    config_sync_platforms = None
    needs_config = (
        args.author is None or args.theme is None or args.sync_from_config
    )
    if not needs_config:
        return config_sync_platforms

    try:
        config = get_config()
    except ConfigError as e:
        if args.sync_from_config:
            print(f"[配置错误] {e}", file=sys.stderr)
            print("--sync-from-config 需要有效的 wechat-publisher.yaml 配置。", file=sys.stderr)
            sys.exit(1)
        if args.author is None:
            args.author = "飞哥"
        return config_sync_platforms

    if args.author is None:
        args.author = config.get("author", "") or "飞哥"
    if args.theme is None:
        args.theme = config.get("theme", "") or None
    config_sync_platforms = config.get("sync_platforms") or None
    return config_sync_platforms


def _resolve_sync_platforms(args, config_sync_platforms):
    """解析最终要同步的平台列表: --sync 显式指定 > 配置里的 sync_platforms。"""
    if args.sync:
        return [p.strip() for p in args.sync.split(",") if p.strip()]
    if args.sync_from_config and config_sync_platforms:
        if isinstance(config_sync_platforms, str):
            return [p.strip() for p in config_sync_platforms.split(",") if p.strip()]
        return [str(p).strip() for p in config_sync_platforms if str(p).strip()]
    return None


def main():
    parser = _build_parser()
    args = parser.parse_args()

    # 设置全局账号
    if args.account:
        set_account(args.account)

    # 从账号配置回读 author / theme / sync_platforms
    config_sync_platforms = _resolve_config(args)

    # 解析最终要同步的平台列表
    sync_platforms = _resolve_sync_platforms(args, config_sync_platforms)

    if not args.exec:
        mode = "newspic" if args.brief or args.type == "newspic" else ("html" if args.html else "news")
        input_path = args.brief or args.html or args.input
        if not input_path:
            parser.error("请提供 --input (Markdown) / --html / --brief(贴图)参数")
        print("dry-run(加 --exec 才会上传素材并创建微信公众号草稿):")
        print(f"  account: {args.account or '(默认账号)'}")
        print(f"  mode:    {mode}")
        print(f"  input:   {input_path}")
        print(f"  title:   {args.title or '(从输入提取)'}")
        return 0

    # --brief 存在 或 --type newspic 走贴图分支
    if args.brief or args.type == "newspic":
        if not args.brief:
            parser.error("--type newspic 需要同时指定 --brief <brief.md>")
        if args.sync or args.sync_from_config:
            parser.error("贴图(newspic)模式不支持多平台同步")
        result = publish_from_brief(
            brief_path=args.brief,
            account_name=args.account,
            image_style=args.image_style,
            title=args.title,
            author=args.author,
            ai_score_threshold=args.ai_score_threshold,
            skip_ai_score=args.skip_ai_score,
        )
    elif args.html:
        if not args.title or not args.cover:
            parser.error("使用 --html 模式时,必须提供 --title 和 --cover")
        if args.sync or args.sync_from_config:
            parser.error(
                "--html 模式不支持多平台同步(HTML 已是微信专用排版)。"
                "如需同步请改用 --input <markdown>。"
            )
        if not args.official_api:
            # 会话式:直接把已排版 HTML 走后台会话建草稿(免 secret/白名单)
            html_content = Path(args.html).read_text(encoding="utf-8")
            result = _session_publish_html(html_content, args.cover, args.title,
                                           args.digest or "", args.author or "",
                                           base_dir=Path(args.html).parent)
        else:
            result = publish_from_html(
                html_path=args.html,
                title=args.title,
                cover_path=args.cover,
                author=args.author,
                digest=args.digest or "",
                source_url=args.source_url,
            )
    elif args.input:
        result = publish_from_markdown(
            md_path=args.input,
            title=args.title,
            author=args.author,
            digest=args.digest,
            cover_path=args.cover,
            source_url=args.source_url,
            temp_dir=args.temp_dir,
            style_path=args.style,
            theme=args.theme,
            sync_platforms=sync_platforms,
            ai_score_threshold=args.ai_score_threshold,
            skip_ai_score=args.skip_ai_score,
            allow_missing_images=args.allow_missing_images,
            session_mode=not args.official_api,
            debug=args.debug,
        )
    else:
        parser.error("请提供 --input (Markdown) / --html / --brief(贴图)参数")

    # 输出JSON结果
    print("\n" + json.dumps(result, ensure_ascii=False, indent=2))


if __name__ == "__main__":
    main()
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall