facebook-research-report
GitHub聚合Facebook研究各阶段数据，生成包含趋势分析、竞品深挖、关键词策略及复制脚本的综合HTML报告。
Trigger Scenarios
Install
npx skills add leamsigc/ShortsGenerator --skill facebook-research-report -g -y
SKILL.md
Frontmatter
{
"name": "facebook-research-report",
"version": "3.0.0",
"description": "Report state — synthesizes all phase outputs into a comprehensive final\nreport. Generates a self-contained HTML report (REPORT.html) plus a markdown\nbackup (report.md). Includes performance summary, history trend comparison,\ntop 5 deconstruction recap, video structure analysis, competitor deep-dive,\nreplication scripts, keyword strategy, and engagement opportunities.\n",
"allowed-tools": [
"Bash",
"Read",
"Write",
"ChromeDevTools",
"WebSearch"
]
}
Facebook Research — Report State (v3)
Generates the final report. The HTML shell is produced by a script; the analytical sections are authored from the scraped data.
Preconditions
- Config:
.fb-research/projects/<project-name>/config.json - Analysis:
.fb-research/projects/<project-name>/top5-analysis.md - Analytics:
.fb-research/projects/<project-name>/page-analytics.json+page-analytics-history.json - Competitors:
.fb-research/projects/<project-name>/competitors/ - Optional video template data:
video-template/VIDEO-TEMPLATE.md,video-analysis.json,transcripts/
Steps
1. Load all data + compute history trends
Read all JSON/markdown files. Compute trend direction per metric (followers, avg views, avg likes, top post views) across snapshots — improving / declining / stable.
2. Discover real competitors via Facebook search
chrome-devtools navigate_page --url "https://www.facebook.com/search/pages/?q=<URL_ENCODED_NICHE>"
Scroll, extract page results, filter to the top 5 most relevant, save to discovered-competitors.json. Benchmark each competitor's positioning relative to the target page.
3. Generate 5 replication script templates
Mix at least 3 from the page's top 5 + up to 2 from competitor patterns. Each script: original source, core pattern, why it works, adapted topic script (hook/body/CTA), caption template, production notes, performance prediction, post-publication tracking checklist.
Save each to replication-scripts/001-<slug>.md and update replication-history.json.
4. Keyword research + competitor deep-dive
- WebSearch the niche for trending topics/audience interests
- Extract keywords from top posts + competitors → keyword clusters by content type
- Visit competitor pages via Chrome DevTools, extract recent posts + video URLs
5. Generate the HTML report + markdown backup
The HTML shell is rendered from the structured data:
node "$CLAUDE_SKILL_ROOT/scripts/report-html.js" <project-name> [sessionId] --session $SESSION
# fallback: node ./skills/facebook-research/scripts/report-html.js <project-name> --session $SESSION
Then augment REPORT.html with the analytical sections authored above (executive
summary, top-5 deep dive, video structure analysis, replication scripts, keyword strategy,
engagement strategy, feedback for page manager, 30-day action plan). Keep REPORT.html
as the primary deliverable and write report.md as the markdown backup. Append a snapshot
to report-history.json.
Report Completion
✅ Facebook Research Report Complete!
Project: .fb-research/projects/<project-name>/
📊 REPORT.html (primary) · 📄 report.md (backup) · 📈 trend-report.md (round 2+)
Replication scripts: <N> · Competitors analyzed: <N>
Next step: skill facebook-research-engage (recommended for growth)
Version History
- 7b83166 Current 2026-08-29 02:30
