fameclaw

name

fameclaw

description

YouTube creator outreach prospecting — find channels by niche, extract emails and stats, discover related channels, and batch-scrape to CSV. Use when asked to find YouTube creators, source creator emails, build influencer/outreach lists, prospect YouTube channels for partnerships, or scrape channel data at scale. Triggers on "find YouTube creators", "source emails", "build outreach list", "find channels in [niche]", "YouTube prospecting", "creator outreach", "scrape YouTube emails", "find influencers".

FameClaw — YouTube Creator Outreach Prospector

Find YouTube creators by niche, extract their emails + stats, discover related channels via recommendations, and batch-scrape everything to CSV.

Zero API keys. Fully local. All data stays on your machine.

Installation

FameClaw scripts live in this skill directory. Set the scripts path:

export FAMECLAW_SCRIPTS="${CLAUDE_SKILL_DIR}/../scripts"
export FAMECLAW_CONFIG_DIR="$HOME/.config/fameclaw"
mkdir -p "$FAMECLAW_CONFIG_DIR"

Requirements: curl, python3 (standard on macOS/Linux). No API keys needed.

Onboarding Flow

For new users, run the onboarding flow before prospecting. Multi-turn conversational flow.

Step 1: Ask for brand info + connect Gmail

Collect: brand name, website URL, Gmail credentials (login, app password, optional send-as alias, display name).

Store at ~/.config/fameclaw/gmail.json (mode 600):

{
  "email": "[e-mail]",
  "app_password": "xxxx xxxx xxxx xxxx",
  "from_email": "[e-mail]",
  "display_name": "Daniel from Brand"
}

Test: python3 $FAMECLAW_SCRIPTS/gmail.py test

Step 2: Scan the site

bash $FAMECLAW_SCRIPTS/onboard.sh --brand "BrandName" --url "https://example.com" --output scan.json

Step 3: Determine product category

From scan, identify category. Categories: Physical products, Digital products, SaaS/App, Services, Marketplace, Creator tools, Info products/Education.

If confident, state it. If unclear, ask one focused question.

Step 4: Propose influencer categories + follower tiers

Based on product category, propose 2-4 YouTube creator types and 2-3 size ranges. See ${CLAUDE_SKILL_DIR}/references/category-mapping.md for the full mapping table.

Step 5: Build audience profile

Create audience.json:

{
  "brand": "BrandName",
  "url": "https://example.com",
  "target_categories": ["Beauty & Skincare", "Fashion & Lifestyle"],
  "target_demographics": { "age_range": "25-40", "gender": "female", "interests": ["skincare"], "location": "US" },
  "authority_preferred": true
}

Ask: "Who is your typical customer?" — age, gender, interests, lifestyle.

Step 6: Generate config

Write config.json with 15-30 tailored search queries, max_subs, target_emails (suggest 100-500).

Step 7: Run prospector

bash $FAMECLAW_SCRIPTS/prospect.sh --config config.json

Step 8: Score and rank

python3 $FAMECLAW_SCRIPTS/score_channels.py --csv channels.csv --profile audience.json --output scored_channels.csv

Adds: content_category, match_score (0-100), match_type, match_reasons. Present top results grouped by match type.

Gmail Outreach

Step 9: Configure outreach

Create outreach.json:

{
  "brand": "MyBrand",
  "website": "https://mybrand.com",
  "sender_name": "Alex",
  "gmail_creds": "gmail_creds.json",
  "current_partnerships": ["@CreatorA", "@CreatorB"],
  "rate": 30,
  "min_score": 25,
  "max_per_run": 50
}

Step 10: Run outreach

python3 $FAMECLAW_SCRIPTS/outreach.py send --csv scored.csv --config outreach.json --dry-run
python3 $FAMECLAW_SCRIPTS/outreach.py send --csv scored.csv --config outreach.json
python3 $FAMECLAW_SCRIPTS/outreach.py check-replies --config outreach.json
python3 $FAMECLAW_SCRIPTS/outreach.py followup --config outreach.json
python3 $FAMECLAW_SCRIPTS/outreach.py status --config outreach.json

Three-stage email sequence, auto-personalized per creator (mentions specific recent videos, current partnerships as social proof, 3-5 sentences max).

Step 11: Negotiate

python3 $FAMECLAW_SCRIPTS/negotiate.py set-config --config outreach.json --key budget_min --value 1000
python3 $FAMECLAW_SCRIPTS/negotiate.py set-config --config outreach.json --key budget_max --value 3000
python3 $FAMECLAW_SCRIPTS/negotiate.py set-config --config outreach.json --key negotiation_style --value value-focused
python3 $FAMECLAW_SCRIPTS/negotiate.py check --config outreach.json
python3 $FAMECLAW_SCRIPTS/negotiate.py status --config outreach.json

Negotiation styles: friendly / value-focused / budget-strict. Full playbook: ${CLAUDE_SKILL_DIR}/references/negotiation_playbook.md

Scheduling (NanoClaw Tasks)

For recurring runs, use NanoClaw's task scheduler instead of OpenClaw cron.

Prospecting (every 30 min until target hit):

Create a task in your NanoClaw config or ask the agent to schedule:

Schedule a recurring task every 30 minutes:
Run: bash $FAMECLAW_SCRIPTS/prospect.sh --config /path/to/config.json
Stop when target email count is reached.

Outreach follow-ups (every 6 hours):

Schedule a recurring task every 6 hours:
Run: python3 $FAMECLAW_SCRIPTS/outreach.py check-replies --config outreach.json && python3 $FAMECLAW_SCRIPTS/outreach.py followup --config outreach.json

Quick Start (advanced users)

# Single channel
bash $FAMECLAW_SCRIPTS/extract_channel_data.sh "https://youtube.com/@handle" output.csv

# Find related channels
bash $FAMECLAW_SCRIPTS/find_related_channels.sh "https://youtube.com/@handle" 20

# Email only
bash $FAMECLAW_SCRIPTS/extract_email.sh "https://youtube.com/@handle"

# Batch prospecting
bash $FAMECLAW_SCRIPTS/prospect.sh \
  --queries "tiktok shop tutorial" "dropshipping beginner" \
  --target 100 --output channels.csv --max-subs 100000

Config JSON

Field	Default	Description
queries	required	YouTube search queries
target_emails	100	Stop after this many emails
output	fameclaw_channels.csv	Output CSV path
max_subs	0 (no limit)	Skip channels above this count
batch_size	200	Max channels per run
work_dir	output dir	State/queue/log directory

CSV Schema

channel_name,handle,subscribers,total_videos,avg_views,median_views,min_views,max_views,videos_sampled,email,description,external_links,channel_url

NanoClaw Container Notes

When running inside a NanoClaw container, ensure these paths are mounted:

• ~/.config/fameclaw/ — credentials (read/write)
• Working directory for CSV output (read/write)
• The FameClaw scripts directory (read-only)

The scripts use only curl and python3 — both available in NanoClaw's default container image.

Limitations

• YouTube hides business emails behind captcha — uses vanity domain workaround
• View counts sampled from visible videos (~30-200)
• Subscriber counts approximate (YouTube rounds)
• Rate limiting: ~0.5s between channels, ~1s between related lookups