Whitelabel Radar

White-Label Radar

Find NM leaders who already sell AI/recruiting education to their own teams, and pitch them a white-labeled Tiger Claw they resell as their own. The model: they teach AI; you give their students the actual agent, branded as theirs. Brent's record selling since 1989 (brentbryson.ai) is the door-opener.

Pattern note: this is a template for last30days-based use-case skills — run the pulse, extract the right entities, segment, act. Clone it for other verticals/intents (recruits, partners, buyers).

Workflow (one todo per step)

1. Pulse — run last30days. Default angle: AI for network marketing and direct selling. Other angles: AI sales agents, network marketing recruiting AI, direct selling AI tools. Take the angle from Brent if he gave one.
   • REQUIRED on this machine: the engine fails with SSL: CERTIFICATE_VERIFY_FAILED unless certifi is wired in. Before running, export it:

     python3.14 -m pip install --user --quiet certifi 2>/dev/null
     export SSL_CERT_FILE="$(python3.14 -c 'import certifi; print(certifi.where())')"
     

     Then run last30days normally (it inherits the env). Without this, every source returns 0 items.
2. Extract LEADERS, not audiences. From the result, pull the named channel owners / creators: YouTube channels, X handles, IG/TikTok creators, book authors. Do NOT harvest commenters - mass-scraping viewers is ToS-gray and against Brent's public-only compliance line. The target is the leader; their audience is reached through the white-label, later.
3. Segment each leader:
   • Tier A — go now: mid-size NM coaches already selling AI/recruiting education (courses, challenges, "how to use AI" content), reachable, real audience. Ideal white-label fit.
   • Tier B — marquee, long game: the big names (e.g. Eric Worre / Network Marketing Pro). Warm intro only (Brent's upline / Jeff Mack), never cold.
   • Competitors — watch, don't pitch: AI-sales-agent products (e.g. CallLoop / @polsia), MLM software vendors. Note them as competitive signal, not partners.
4. Find the PUBLIC contact path for each Tier A/B target via Codex web search — one-by-one, public only: website contact/booking page, YouTube "About" business email, LinkedIn, IG DM. Never a scraped blast.
5. Write the dossier to ~/Desktop/whitelabel-target-dossier.md (update if it exists): a table per tier (name · where · what they sell · contact path · tailored angle), plus a leader-to-leader outreach template that opens with genuine reference to their work, states Brent's top-rank credibility, links brentbryson.ai, and offers a 15-minute white-label demo. Value-first, no grind.
6. Update the pipeline tracker at ~/Desktop/partner-pipeline.md: append any genuinely-new target as a new row (target · tier · channel · status · last touch · next action · notes). NEVER overwrite an existing row's status - only add new ones and append a dated line to the Activity log. Then hand Brent the action the frictionless way — he is phone-first and hates copy-paste / markdown:
   • Emails (the default): queue a FINISHED draft straight into his Gmail Drafts. Write the body (message + signature) to a temp file, then run: python3 ~/.tiger/draft_email.py --to "<addr>" --subject "<subj>" --body-file <tmp> It appears in his phone Gmail → Drafts → he taps Send. NO paste-blocks, NO Reminders for emails. Batch it: for a multi-target run, queue ALL the drafts in one pass so he sends a stack, not one-at-a-time. (Mechanism + app-password handling: see memory project_gmail_draft_pipeline.) Only fall back to a paste-ready To/Subject/Body block in chat if the draft pipeline is unavailable.
   • Non-email actions (a call, a decision, a real-world to-do): add to the Apple "DO THIS" Reminders list via osascript (syncs to his iPhone, voice-addable). Do NOT use ~/Desktop/DO-THIS.md markdown — he never opens markdown.
   • Never let a manual step live only in chat.
7. Hand off to tiger-leader-hunt for deep scoring/enrichment of the Tier A/B list (leverage, team size, reach). This skill builds the seed list; tiger-leader-hunt ranks and deepens it.

Outreach template (leader-to-leader, no grind)

[Name] - I saw your [specific video/post] on AI for [recruiting/lead-gen]. You're teaching the right thing. I'm Brent Bryson - selling since 1989, top rank in network marketing - and I built the agent your students are looking for: it does the follow-up, books the call, keeps every lead warm. Here's me: brentbryson.ai. I'd love to show you a white-labeled version you could put your name on for your team. Worth 15 minutes?

Out of scope (point elsewhere)

• "Show up as the top voice" is a content track, not this skill - use cadence / write-content, and genuinely engage on these leaders' posts before pitching.
• Deep per-leader scoring → tiger-leader-hunt. Trend research → last30days.

Codex Runtime

Resolve sibling Codex skills from installed skill roots (including ~/.codex/skills) or the current collection; do not assume a legacy skill root.

Never expose or print secret, credential, or token values.

Mandatory dependencies:

• last30days and tiger-leader-hunt skills
• enrichment and publishing credentials

Preflight each dependency using MCP/app capability discovery, CLI availability/version checks, read-only filesystem or Git checks for repositories, and provider auth-status commands without printing secrets, credentials, or tokens. If any mandatory dependency is unavailable, stop and report a concise blocked state naming the missing dependency and the next action needed.