Turning public signals into a qualified first customer shortlist

What it does

This Codex skill takes a startup URL or product description and produces a ranked shortlist of potential first customers. It researches recent public signals—explicit complaints, workarounds, switching intent, and timing triggers—then links every prospect back to the original source. The final output is a standalone HTML report with fit scores, signal dates, and draft outreach openers, but it deliberately stops short of sending anything.

The interesting bit

Rather than scraping private contact databases or buying enrichment data, it treats public posts, discussions, and business triggers as primary research. The skill builds an evidence folder for each prospect before you write a single sentence, which is exactly the step most founders skip.

Key highlights

• Evidence-based scoring: every prospect carries a fit and timing score tied to a public source link and date
• Multiple research modes: quick (5 prospects), standard (10), deep (20), plus design-partner, B2B, and community-specific searches
• Built-in guardrails: no automated sending, no private contact enrichment, and no sensitive personal data handling
• Deliverable is a responsive HTML report with ICP definition, repeated pain patterns, and a 7-day manual outreach plan
• Designed for the Codex CLI ecosystem; installs as a skill under ~/.codex/skills/

Caveats

• The README explicitly warns that prospects are “hypotheses based on public signals, not confirmed customers or guaranteed buyers,” so the output is starting material, not a sales guarantee.

Verdict

Best for early-stage founders who need to turn a product idea into a concrete outreach list without hiring a sales researcher. If your pipeline is already full, skip it.