{"skill":{"slug":"kol-content-screening","displayName":"Kol Content Screening","summary":"Screen and rank Chinese social media KOLs by matching keyword content within a time window using web search aggregation, reporting evidence and confidence.","description":"---\nname: kol-content-screening\ndescription: Screen and rank Chinese social media KOLs (抖音 Douyin / 小红书 Xiaohongshu / 今日头条 Toutiao / 视频号 / B站) by whether they have published content matching a given keyword (brand, competitor, topic) within a time window, using web search aggregation rather than platform APIs. Use when the user provides a list of KOL accounts (handle / nickname / fan count / homepage URL) and asks to find which ones posted about X (e.g., 比亚迪 / 宁德时代 / DM-i / 王传福) in the past N months, or to rank/sort them by relevance / fan count / interaction. Triggers on phrases like \"筛选达人\", \"KOL 内容查询\", \"媒体账号比亚迪相关内容\", \"查这批博主有没有发过 X\", \"投放筛选\". Optimized for marketing / PR / competitive-intel workflows where the platform's own API is unavailable. NOT for: real per-video interaction counts (need 蝉妈妈/灰豚/新红 paid data services), discovering new accounts from scratch, or non-Chinese platforms.\n---\n\n# KOL Content Screening (Web-Search Based)\n\nScreen Chinese social media KOL lists for keyword-matching content within a time window. Output a ranked, evidence-backed report. Used heavily in PR / marketing / competitive intel work where you receive a \"已知账号清单 + 关键词 + 时间窗\" and need to know \"谁发过、谁没发\"。\n\n## Hard Truths Up Front\n\nTell the user these before promising anything:\n\n1. **No reliable per-video interaction counts via web search.** 抖音/小红书 single-video 点赞/评论/收藏 are not stably indexed by general search engines. Mark as \"无公开数据\" rather than guess. For real numbers the user must use 蝉妈妈 / 灰豚 / 新红 (抖音), 千瓜 / 新红 (小红书), or platform open APIs.\n2. **\"未发现\" ≠ \"没发过\".** Web search has indexing gaps. Always frame negatives as \"公开检索未发现证据 (within N months)\". Never claim a creator definitely hasn't posted X.\n3. **Same nickname ≠ same person.** Verify by handle/UID/homepage URL, not by name. 抖音号 / 小红书 user_id / 头条 author UID are the only reliable identifiers.\n4. **Time window matters.** State the explicit window (e.g. `2025-05-05 ~ 2026-05-05`) in the report header. Old content (>1 year) gets marked separately, not mixed into \"active\" set.\n\nIf the user asks for accurate single-video互动量 排序, **stop and warn**: this needs paid data services. Get explicit acknowledgement before proceeding with web-only screening.\n\n## Core Workflow\n\n### 1. Intake (clarify before running)\n\nAlways confirm 5 parameters before spawning sub-agents:\n\n| Parameter | Example | Notes |\n|-----------|---------|-------|\n| Platforms | 抖音 + 小红书 + 头条 | Each platform = independent sub-task |\n| Account list | (CSV/table from user) | Need: handle/UID + nickname + fan count + homepage URL |\n| Keywords | 比亚迪 / BYD / 王传福 / DM-i / 仰望 | Include EN + CN + product lines + key person names |\n| Time window | 近 12 个月 (`YYYY-MM-DD ~ YYYY-MM-DD`) | Compute exact dates; don't pass \"近一年\" verbatim |\n| Sort dimension | 有内容档→粉丝量降序 / 互动量 / 关键词命中数 | Without 互动量数据来源, default to fan count desc within match-tier |\n\n**Common intake mistake**: User pastes a Windows-clipboard HTML fragment (`Version:1.0 StartHTML:...`) — that's the raw clipboard envelope. The actual table data is below it. Parse account list directly from the rest of the paste.\n\n### 2. Group & Parallelize\n\nFor >15 accounts on one platform, split into groups of 8–10 and spawn parallel sub-agents. Empirically: 36 抖音 accounts → 4 groups of ~9, 24 小红书 accounts → 3 groups of 8.\n\n```\nPer platform → 拆 N 组 → 每组 1 sub-agent → 并行 → 各自写文件 → 主 session 汇总排序\n```\n\nEach sub-agent writes ONE file. File naming convention:\n```\n{platform-prefix}-{keyword-slug}-research-group{N}.md\n```\nwhere `platform-prefix` is `douyin` / `xhs` / `tt` / `sph` (视频号) / `bilibili`.\n\nSub-agent prompt template: see `references/subagent-prompt-template.md`.\n\n### 3. Per-account search procedure\n\nEach sub-agent, for each account, runs **at least two** queries on the chosen web search tool (xiaosu-search or equivalent):\n\n```\nQ1: \"{nickname}\" {handle} {keyword}\nQ2: \"{nickname}\" {keyword} site:{platform-domain}\nQ3 (if Q1+Q2 weak): {nickname} {keyword} {YYYY}   # last 12 months explicit\n```\n\nWhere `{platform-domain}` is `douyin.com` / `xiaohongshu.com` / `toutiao.com` / etc.\n\nFor each hit, the sub-agent records:\n- **Title** (or first line of post)\n- **Date** (verify within window — outside-window hits noted separately)\n- **URL** (must point to the creator's own post, not third-party reposts/quotes)\n- **Stance** (正向 / 中性 / 负向 / 仅提及) — affects PR usability\n- **Confidence** (高 / 中 / 低) — based on evidence strength\n\nFor each account, the sub-agent must explicitly check for **ID collision**: search for the nickname alone, see if the top hits are this person's handle. If collision is detected (e.g. \"南希Nancy\" — multiple persons), flag it.\n\nSee `references/platform-search-tips.md` for platform-specific quirks (site filters, profile URL formats, common false positives).\n\n### 4. Aggregate & Rank\n\nMain session reads all group files and merges into one ranked table. Default ranking:\n\n```\nTier 1 🟢  — 近一年内有明确证据（带 URL、日期、内容摘要）\nTier 2 🟡  — 仅旧内容（>窗口）/ 间接提及 / 证据较弱\nTier 3 🔴  — 公开检索未发现\n```\n\nWithin each tier: sort by fan count desc by default. If user asked for interaction-based ranking but data is unavailable, **state this explicitly** in the report and fall back to fan count + provide caveat.\n\nFinal report structure: see `references/output-schema.md`.\n\n### 5. Deliver\n\nOutput to `<workdir>/<keyword-slug>-kol-screening-{YYYYMMDD}.md` (markdown table) plus per-platform group files. If user wants 飞书 Sheet, build the markdown first, then offer to push via `lark-cli sheets` (separate skill).\n\n## Failure Modes Seen In The Wild\n\nDocument these in the report so the user can interpret correctly:\n\n- **Handle drift** — User pastes \"楠姐财经科技头条\" but real similar accounts are \"楠姐聊财经\" / \"楠姐科技说\" / \"楠姐谈股论今\". Report all candidates, flag uncertainty, ask user to confirm.\n- **Cross-platform leak** — A 视频号 creator's content shows up only via 新浪/百度 reposts. That's still valid evidence the post exists, but mark source as `via 新浪 (转载)`.\n- **Brand homonyms** — \"比亚迪\" matches food brand 拿铁/拿铁酱 etc.; \"宁德时代\" rarely collides but \"宁德\" alone matches geography. Use full brand name + a disambiguator keyword (王传福, 刀片电池, 车型名 for BYD; 麒麟电池, 神行, 凝聚态 for CATL).\n- **Stale fan counts** — User-supplied fan numbers are snapshots. Don't recompute; record as-given with date if user provided one.\n- **Profile-not-found** — Sometimes the homepage URL in the user's list 404s. Report as \"主页失效\", do NOT skip the account silently.\n\n## Honest Reporting Discipline\n\nEvery report MUST include a **methodology disclaimer block** at the top:\n- Data source (web search, which provider)\n- What CAN'T be obtained (per-video interaction counts, follower-only content, etc.)\n- Time window (explicit dates)\n- Confidence framing (\"未发现 ≠ 没发过\")\n\nTemplate in `references/output-schema.md`.\n\n## Quick Reference\n\n- Sub-agent prompt template → `references/subagent-prompt-template.md`\n- Platform-specific search tips → `references/platform-search-tips.md`\n- Output schema + methodology block → `references/output-schema.md`\n- Decision table: when to refuse / when to upgrade to paid data → `references/escalation.md`\n","topics":["Marketing","Web Search","小红书","抖音"],"tags":{"byd":"0.1.0","catl":"0.1.0","chinese-social-media":"0.1.0","kol":"0.1.0","latest":"0.1.0","marketing":"0.1.0","pr":"0.1.0","web-search":"0.1.0"},"stats":{"comments":0,"downloads":339,"installsAllTime":12,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1778123216462,"updatedAt":1778492867743},"latestVersion":{"version":"0.1.0","createdAt":1778123216462,"changelog":"Initial release: web-search-based KOL content screening for 抖音/小红书/头条/视频号/B站. Distilled from real CATL project work screening 70 KOLs across 3 platforms for BYD content (2026-05-05).","license":"MIT-0"},"metadata":null,"owner":{"handle":"dr-xiaoming","userId":"s17bgyae9q08vjv1yjyeatsq6h83jkzj","displayName":"Dr-xiaoming","image":"https://avatars.githubusercontent.com/u/88608717?v=4"},"moderation":null}