content-growth-review
Skill14 starsChanged 5 months agoMITView on GitHub
Analyze updated creator media data after incremental crawls and turn it into content growth diagnosis, direction adjustments, and complete topic planning. Use when the user asks to复盘抖音/视频号/公众号/小红书数据, 分析为什么涨粉弱或互动异常, 做内容增长计划, 从评论生成选题, 更新 raw/studio/funnel/backlog/manual topic pools, or plan the next publishing direction after running kangjian-douyin-media or other platform archives.
In this repository39 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- logo-generatorSkill
- source-command-kj-helpSkill
- ai-coding-weekly-reportSkill
- ai-conversation-to-wikiSkill
- ai-daily-from-xSkill
- ai-daily-websearchSkill
- ai-learning-loopSkill
- ai-programming-topic-plannerSkill
- ai-reading-coachSkill
- coding-session-to-tutorialSkill
- content-creatorSkill
- content-illustratorSkill
- course-campaignSkill
- course-generatorSkill
- course-scraperSkill
- english-reading-studioSkill
- extract-conversation-insightsSkill
- git-pushSkill
- humanizer-zhSkill
- industry-best-practicesSkill
- kangjian-douyin-mediaSkill
- kangjian-skillSkill
- knowledge-mapSkill
- markdown-formatSkill
- paid-content-reviewSkill
- phenomenon-insightSkill
- research-backed-content-loopSkill
- skills-repo-health-checkSkill
- story-pptSkill
- transcript-critical-analysisSkill
- tutorial-guideSkill
- web-slideSkill
- wechat-companionSkill
- wechat-publish-kitSkill
- weekly-digestSkill
- wiki-doc-sinkSkill
- wiki-intelligenceSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
20 rules. About 0 are common, 9 are distinctive, and 11 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
Distinctive · estimate
This author's own choices. Worth a look.
11
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Use this skill after media data has been updated.
- Very specific: line 10
Use $kangjian-douyin-media or another platform collector first to refresh raw data.
- Distinctive: line 11
Use this skill to diagnose growth problems, decide content direction, and produce topic plans.
- Distinctive: line 12
Default workspace is the LLM Wiki repo.
- Very specific: line 13
Default review output path: raw/studio/funnel/reviews/.
- Very specific: line 14
Default topic output path: raw/studio/funnel/backlog/manual/.
- Distinctive: line 19
Read the generated input bundle and any referenced source reports.
- Very specific: line 30
Write selected topic pools to raw/studio/funnel/backlog/manual/ unless the user only wants chat advice.
- Distinctive: line 45
Always check the relevant files if they exist:
- Distinctive: line 61
Write a dated Markdown report when doing a real review:
- Distinctive: line 82
Use concise tables for topic suggestions.
- Very specific: line 96
When updating raw/studio/funnel/backlog/manual/, create one dated topic-pool file instead of scattering many tiny files:
- Distinctive: line 102
Use frontmatter compatible with the local manual backlog:
- Very specific: line 121
Do not edit queue.md unless the user explicitly asks to queue items for this week.
- Very specific: line 156
Prefer using AI 编程产品实战 as the mother theme.
- Very specific: line 160
Use $kangjian-douyin-media first when the user asks to update Douyin data.
- Very specific: line 161
Use $content-growth-review after data exists and the user asks for复盘、增长计划、内容调整、选题建议.
- Very specific: line 162
Use $local-stt-transcription when review conclusions require missing video transcripts.
- Distinctive: line 169
Run git diff --check on changed review/topic files.
- Very specific: line 171
If queue.md was not edited, say that topics remain in status: idea.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Content Growth Review
Use this skill after media data has been updated. It is the analysis layer, not the crawler.
- Use
$kangjian-douyin-mediaor another platform collector first to refresh raw data. - Use this skill to diagnose growth problems, decide content direction, and produce topic plans.
- Default workspace is the LLM Wiki repo.
- Default review output path:
raw/studio/funnel/reviews/. - Default topic output path:
raw/studio/funnel/backlog/manual/.
Core Workflow
- Collect the latest inputs with
scripts/collect_growth_inputs.py. - Read the generated input bundle and any referenced source reports.
- Diagnose data problems by separating:
- content-market fit
- topic clarity
- account positioning
- hook/opening strength
- format and series continuity
- comment-to-follow conversion
- product/course/community conversion
- Produce a review report.
- Convert high-value findings into topic suggestions.
- Write selected topic pools to
raw/studio/funnel/backlog/manual/unless the user only wants chat advice.
Quick Start
python3 <skill_dir>/scripts/collect_growth_inputs.py \
--wiki-root "<wiki-root>" \
--platform douyin \
--output "/tmp/content-growth-review-inputs.md"Then read /tmp/content-growth-review-inputs.md and the linked source files.
Required Reads
Always check the relevant files if they exist:
raw/studio/funnel/published/<platform>/latest-*.md- latest
raw/studio/funnel/published/<platform>/*review-analysis.md - latest
raw/studio/funnel/published/<platform>/*full-ledger.mdonly when comment examples or exact work details are needed raw/studio/funnel/backlog/manual/raw/studio/funnel/backlog/queue.mdwiki/playbooks/topic-bank/series-planning.md
For Douyin, also check:
raw/studio/funnel/published/douyin/manifest/*.jsonif presentraw/studio/funnel/published/douyin/runs/when comparing snapshots
Review Report Contract
Write a dated Markdown report when doing a real review:
raw/studio/funnel/reviews/YYYY-MM-DD-<platform>-content-growth-review.mdInclude these sections:
# <平台>内容增长复盘## 数据范围## 本次最重要结论## 数据异常与问题诊断## 内容支柱表现## 账号定位与关注理由## 评论洞察## 内容调整建议## 后续方向规划## 完整选题建议## 本周建议晋升## 下次增量更新要观察什么
Use concise tables for topic suggestions. Every recommended topic should include:
- topic/title
- target audience
- source signal
- published connection
- series position
- expected metric
- first hook
- next episode hook
- whether it should enter
manual/orqueue.md
Topic Pool Contract
When updating raw/studio/funnel/backlog/manual/, create one dated topic-pool file instead of scattering many tiny files:
raw/studio/funnel/backlog/manual/YYYY-MM-DD-<platform>-growth-topics.mdUse frontmatter compatible with the local manual backlog:
---
title: "..."
source: manual
pillar: ai-coding
platform: shipin
kind: funnel
status: idea
lesson_id: ~
snippet_link: ~
formatted_link: ~
published_link: ~
priority: 1
notes: "..."
---Do not edit queue.md unless the user explicitly asks to queue items for this week. If queueing, keep the number small: 1-3 items.
Analysis Heuristics
Interpret metrics by intent:
- High saves: utility, tutorial value, future use.
- High comments: controversy, confusion, unmet need, identity tension.
- High shares: social proof or useful decision aid.
- High likes but low saves/comments: entertaining or agreeable, weaker conversion.
- High comments but low saves: debate/情绪; useful for attention but may not convert.
- High saves but low comments: tutorial/reference; useful for course and资料包 conversion.
For growth diagnosis, distinguish:
曝光问题: topic/hook/distribution may be weak.互动问题: claim is not concrete or not emotionally loaded.关注问题: account promise is unclear or content types feel scattered.转化问题: CTA,资料包,群聊,课程 promise is not connected to the video topic.系列问题: strong videos do not naturally tell users what comes next.
Kangjian Positioning
Current strategic direction:
帮想用 AI 编程的人,把想法真正做成产品。
Audience is broader than programmers:
- programmers improving AI coding productivity
- non-technical vibe coding users
- independent makers and small founders
- product/operations people building MVPs
- creators turning workflows into products
Prefer using AI 编程产品实战 as the mother theme. Treat AI 编程教程实战 as the method library and 自媒体内容创作工作流 as proof/case material.
Relationship To Other Skills
- Use
$kangjian-douyin-mediafirst when the user asks to update Douyin data. - Use
$content-growth-reviewafter data exists and the user asks for复盘、增长计划、内容调整、选题建议. - Use
$local-stt-transcriptionwhen review conclusions require missing video transcripts.
Validation
Before final response:
- Confirm the reviewed source files and platform.
- Run
git diff --checkon changed review/topic files. - Report every file created or updated.
- If
queue.mdwas not edited, say that topics remain instatus: idea.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

