client-research

Client Research Skill

Synthesize multiple research sources into a comprehensive client strategy document optimized for both human review and downstream AI agents (GTM strategy, list building, email copywriting).

Required Context

Before generating the research document, load all available client sources. {workspace} = the client campaign workspace (default ./cold-email/{slug}/ under the current directory):

Directory Structure Reference:

{workspace}/
├── research.md       # Existing research (preserve metadata if updating) - THIS skill's output
├── notes.md          # Optional client preferences
├── research/         # Source material: scraped website content,
│   └── website/      # transcripts, documents the user dropped in
└── {plan folders}/   # Past campaign plans (note what worked)

Sources may also arrive as user uploads or paths given in the conversation - use whatever is provided.

Inputs

Expect one or more of these source types:

• Meeting recordings/transcripts (Fireflies, Otter, Gong, etc.)
• Scraped website content
• Product documentation
• Competitor analysis
• Customer testimonials or case studies
• Sales materials

Output Format

Generate ONE document, {workspace}/research.md, with a quick-reference summary on top followed by the comprehensive research. This single file is what every downstream campaign skill reads.

Part 1: Summary (top of research.md)

Quick-reference section for daily use. See references/context-template.md for the exact structure. Key sections:

• Header metadata (client slug, created date, owner)
• Elevator Pitch - Short hook answering: "If I met {ICP} in an elevator with {Company}, what should I say in a couple seconds to make them really hooked?"
• About the Company - 3-paragraph executive summary
• Target Audience - Condensed ICP (1-2 paragraphs)
• Core Problems Solved - Bulleted list (3-5 items)
• Key Assets - Top 3 lead magnets (table format)
• Case Studies - Top 3 results (table format)
• Quick Links - Website, other client links