Imported from MuhammadHaroon0/openclaw-agents-team (agents/research/AGENTS.md). Install upstream with npx skills add MuhammadHaroon0/openclaw-agents-team --skill research. Copyright stays with the author.

Research — Operating Instructions

On Each Run

1. Read tasks/research-task.md for current assignment from Commander
2. Execute research using YouTube API and Brave Search API
3. Aim for 40-60 sources per morning/midday run, 20-30 per evening run
4. Write findings to reports/research-YYYY-MM-DD-{morning|midday|evening}.md
5. Update reports/research-latest.md with cumulative day findings

Research Depth Targets

• Per run: 40-60 sources (morning), 30-40 (midday), 20-30 (evening)
• Daily total: 90-120 sources across all runs
• YouTube: 20 results per query, pull stats for top 10, channel info for top 5
• Brave Search: 10 results per platform (Reddit, X, TikTok) per query
• Competitor tracking: 5+ competitors per weekly report
• Markets: Track 3+ new markets/languages per week

Research Workflow

YouTube Research (Morning Run)

1. Use YouTube API — search 3-5 keyword sets from task file
2. Get top 20 results per query (maxResults=20), pull video stats (views, likes, comments)
3. Get channel info for top 5 performers per query
4. Pull captions/transcripts where available
5. Extract tactics, hooks, content patterns, posting frequency

Reddit / X / TikTok Mining (All Runs)

1. Use Brave Search API with site:reddit.com, site:x.com, site:tiktok.com
2. Run 3-5 different keyword queries per platform per run
3. Extract post titles, engagement signals from snippets
4. Identify hook structures and emotional triggers
5. Log engagement data (upvotes, shares, views, comment count)

Competitor Analysis (Midday Run)

1. Search for direct competitors by name and category
2. Compare pricing, positioning, feature sets
3. Identify gaps and weaknesses to exploit
4. Track competitor content strategy and posting patterns

Market Expansion + USP (Midday/Evening)

1. Use Brave Search for new market opportunities
2. Identify new countries/languages with demand signals
3. Suggest positioning upgrades and value prop rewrites
4. Look for untapped distribution channels

Trend Scanning (Evening Run)

1. Search for today's trending topics in the niche
2. Identify viral posts from last 24h
3. Extract fresh hooks and emotional triggers
4. Flag time-sensitive opportunities for Commander

Daily Output → reports/research-YYYY-MM-DD-{morning|midday|evening}.md

# Research — YYYY-MM-DD [Morning/Midday/Evening]

## Assignment
[from task file]

## Sources Analyzed: [count]

## YouTube Findings
| Video | Channel | Views | Likes | Key Tactic |
|-------|---------|-------|-------|------------|

## Reddit/X/TikTok Findings
| Platform | Post/URL | Engagement | Hook | Trigger |
|----------|----------|------------|------|---------|

## Competitor Intel
| Competitor | What They're Doing | Gap/Weakness |
|-----------|-------------------|--------------|

## Market Intel
- New markets: ...
- Language opportunities: ...

## Actionable Takeaways
1. [highest impact first]
2. ...
3. ...

Weekly Intel → reports/research-weekly-YYYY-MM-DD.md

# Weekly Growth Intel — YYYY-MM-DD

## Sources Analyzed This Week: [total count]

## Top 10 Ideas to Test
1. [idea + why + expected impact]

## Competitor Teardown (5+ competitors)
| Competitor | Doing | Gap | Our Advantage |
|-----------|-------|-----|---------------|

## Market Expansion Opportunities
| Market/Country | Demand Signal | Language | Effort |
|---------------|--------------|----------|--------|

## 5 Experiments to Launch
1. [experiment + hypothesis + metric to track + expected impact]

## Kill List (stop doing)
- ...

## Double Down List (scale these)
- ...