mcpField
B
baokuan-article-analysis

baokuan-article-analysis

github.com/spacezephyrMarketing & growth1.8k60 installsUpdated 9w ago
Repository

About baokuan-article-analysis

baokuan-article-analysis is a skill for sector- or keyword-level analysis of hot WeChat Official Account articles. Its script merges and deduplicates article data and produces an HTML report; it is not intended to retrieve one account’s complete history.

Key features

  1. 01Fetches hot articles by sector or keyword
  2. 02Merges keyword results and ranks deduplicated items
  3. 03Writes raw JSON data
  4. 04Generates an HTML report with charts and analysis
  5. 05Examines reads, shares, and writing patterns

Use cases

  1. 01Track recent WeChat articles in a sector
  2. 02Compare popular articles across keywords
  3. 03Find writing patterns in highly read or shared posts
  4. 04Inform content topics and writing

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/spacezephyr/creator-buddy --skill baokuan-article-analysis

After installing, type/baokuan-article-analysisin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

agents/
openai.yaml
references/
default-sectors.json
scripts/
daily_sector_trends.py

Details

From the project README

爆款文章分析

Overview

Use this skill to fetch hot WeChat Official Account article data by sector and generate a daily analysis report. It is for sector-level or keyword-level analysis, not exact historical scraping for one specific account.

The bundled script queries the hot-article data source, merges keywords by sector, deduplicates articles, ranks them, and writes:

  • data.json: raw structured data
  • report.html: a minimal visual analysis report with KPI cards, bar charts, ranked article cards, writing style analysis, hot reasons, and writing references

Do not default to Markdown reports. The primary user-facing artifact is report.html.

Quick Start

Run with the default sectors:

python3 ~/.codex/skills/baokuan-article-analysis/scripts/daily_sector_trends.py \
  --output-dir ./output/baokuan-article-analysis

Run custom sectors:

python3 ~/.codex/skills/baokuan-article-analysis/scripts/daily_sector_trends.py \
  --sector 'AI Agent=AI Agent,智能体,Agent框架' \
  --sector 'Skill=skill,Skills,AI Skill' \
  --sector 'Claude Code=Claude Code,Codex,AI编程' \
  --output-dir ./output/baokuan-article-analysis

Run with a JSON config:

python3 ~/.codex/skills/baokuan-article-analysis/scripts/daily_sector_trends.py \
  --sector-config ~/.codex/skills/baokuan-article-analysis/references/default-sectors.json \
  --days 7 \
  --output-dir ./output/baokuan-article-analysis

Workflow

  1. Identify sectors and keywords from the user request.
  2. If the user only gives broad sectors, use or adapt references/default-sectors.json.
  3. Run scripts/daily_sector_trends.py.
  4. Open the generated report.html.
  5. Summarize for the user:
    • highest-reading and highest-sharing articles
    • writing style patterns
    • hot article reasons
    • title and topic formulas
    • practical writing references
  6. Return the HTML file path as the main artifact.

Script Options

OptionPurpose
--sector '赛道=关键词1,关键词2'Add one sector. Can repeat.
--sector-config path.jsonLoad sectors from JSON object.
--days NLookback window. Default is 7 days.
--start-date YYYY-MM-DDExplicit start date. Overrides --days.
--max-items-per-sector NLimit ranked articles per sector. Default is 10.
--output-dir DIRParent output directory. A date folder is created inside.
--report-date YYYY-MM-DDReport date. Default is today.

Data Boundaries

  • This skill returns hot-list data, not a specific account’s complete recent history.
  • clicksCount is a public data-source snapshot and may lag behind live WeChat backend reads.
  • If a specific account name returns no data, switch to that account’s topic keywords and compare same-sector articles.
  • If today’s data is sparse, use --days 7 or --days 30.

Output Interpretation

Use reading count for reach, share count for spread, comments for discussion, and low-fan high-reading entries for title/structure references. Repeated accounts indicate strong competitors or content sources worth following.

Comments

Markdown supported