agentleFS
chengkj99chengkj99/kj-skills/skills/content-growth-review/SKILL.md

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.

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
  1. Distinctive:

    Use this skill after media data has been updated.

    line 8
  2. Very specific:

    Use $kangjian-douyin-media or another platform collector first to refresh raw data.

    line 10
  3. Distinctive:

    Use this skill to diagnose growth problems, decide content direction, and produce topic plans.

    line 11
  4. Distinctive:

    Default workspace is the LLM Wiki repo.

    line 12
  5. Very specific:

    Default review output path: raw/studio/funnel/reviews/.

    line 13
  6. Very specific:

    Default topic output path: raw/studio/funnel/backlog/manual/.

    line 14
  7. Distinctive:

    Read the generated input bundle and any referenced source reports.

    line 19
  8. Very specific:

    Write selected topic pools to raw/studio/funnel/backlog/manual/ unless the user only wants chat advice.

    line 30
  9. Distinctive:

    Always check the relevant files if they exist:

    line 45
  10. Distinctive:

    Write a dated Markdown report when doing a real review:

    line 61
  11. Distinctive:

    Use concise tables for topic suggestions.

    line 82
  12. Very specific:

    When updating raw/studio/funnel/backlog/manual/, create one dated topic-pool file instead of scattering many tiny files:

    line 96
  13. Distinctive:

    Use frontmatter compatible with the local manual backlog:

    line 102
  14. Very specific:

    Do not edit queue.md unless the user explicitly asks to queue items for this week.

    line 121
  15. Very specific:

    Prefer using AI 编程产品实战 as the mother theme.

    line 156
  16. Very specific:

    Use $kangjian-douyin-media first when the user asks to update Douyin data.

    line 160
  17. Very specific:

    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.

    line 161
  18. Very specific:

    Use $local-stt-transcription when review conclusions require missing video transcripts.

    line 162
  19. Distinctive:

    Run git diff --check on changed review/topic files.

    line 169
  20. Very specific:

    If queue.md was not edited, say that topics remain in status: idea.

    line 171

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Content Growth Review

Use this skill after media data has been updated. It is the analysis layer, not the crawler.

  • Use $kangjian-douyin-media or 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

  1. Collect the latest inputs with scripts/collect_growth_inputs.py.
  2. Read the generated input bundle and any referenced source reports.
  3. 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
  4. Produce a review report.
  5. Convert high-value findings into topic suggestions.
  6. 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.md only when comment examples or exact work details are needed
  • raw/studio/funnel/backlog/manual/
  • raw/studio/funnel/backlog/queue.md
  • wiki/playbooks/topic-bank/series-planning.md

For Douyin, also check:

  • raw/studio/funnel/published/douyin/manifest/*.json if present
  • raw/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.md

Include these sections:

  1. # <平台>内容增长复盘
  2. ## 数据范围
  3. ## 本次最重要结论
  4. ## 数据异常与问题诊断
  5. ## 内容支柱表现
  6. ## 账号定位与关注理由
  7. ## 评论洞察
  8. ## 内容调整建议
  9. ## 后续方向规划
  10. ## 完整选题建议
  11. ## 本周建议晋升
  12. ## 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/ or queue.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.md

Use 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-media first when the user asks to update Douyin data.
  • 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.
  • Use $local-stt-transcription when review conclusions require missing video transcripts.

Validation

Before final response:

  1. Confirm the reviewed source files and platform.
  2. Run git diff --check on changed review/topic files.
  3. Report every file created or updated.
  4. If queue.md was not edited, say that topics remain in status: idea.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.