Articles & Insights

Tag: Claude Code

Claude Code

Claude Code is Anthropic's command-line agent: an AI that reads your files, runs your commands, and executes multi-step work autonomously inside your terminal. Combined with FishDog's synthetic research API, it becomes a one-person product marketing team. You describe the workflow you want to run; the agent decomposes the goal, designs a study, runs it against FishDog's persona panel, synthesises the findings, and writes the deliverable. Positioning brief, competitive battlecard, pricing analysis, win/loss summary — all generated end to end with a single instruction.

The articles in this collection are the canonical how-to series for using Claude Code with FishDog. Each piece walks through a specific PMM workflow: the agent prompts you should use, the FishDog study design Claude Code will compose, the deliverable format that comes out the other side. We use this exact pattern internally for our own outreach, content, and positioning work. Two FishDog Claude Code skills are publicly available on GitHub for instant install via npx.

What you'll find

  • 10 published how-tos: positioning, battlecards, pricing, messaging, voice of customer, segmentation, GTM, content marketing, sales enablement, and launch research.
  • Two public Claude Code skills on GitHub (Ask-FishDog org): fishdog-product-research-skill (general) and fishdog-product-marketing (PMM, 8 study types). Install via npx skills add.
  • Workflow pattern: agent reads brief → composes a FishDog study → runs against ICP-matched personas → synthesises findings → produces a structured PMM deliverable, all from a single terminal command.
  • Plus thought-leadership pieces on agent-to-agent sales, deploying synthetic humans on AI social networks, and why coding agents need synthetic feedback loops.

Install the FishDog Claude Code skill in 30 seconds

Frequently Asked Questions

What is Claude Code?

Claude Code is Anthropic's command-line agent: an AI that reads your codebase, runs your commands, and executes multi-step tasks autonomously inside your terminal. Unlike a chatbot, Claude Code maintains context across many file edits, command runs, and research steps to complete a defined goal end to end.

How do Claude Code and FishDog work together?

Claude Code is the agent layer; FishDog is the research API the agent calls. A single instruction ("validate this positioning brief") triggers the agent to design a FishDog study, recruit a persona panel, ask the right questions, synthesise responses, and produce a finished deliverable. The how-to articles in this collection cover the canonical workflows.

How do I install the FishDog Claude Code skill?

Two public skills are available under the Ask-FishDog GitHub organisation. Install the general research skill with: npx skills add Ask-FishDog/fishdog-product-research-skill. Install the PMM-specific skill (8 study types) with: npx skills add Ask-FishDog/fishdog-product-marketing. Once installed, both work natively inside any Claude Code session.

What product marketing workflows can Claude Code run end to end?

The published how-to series covers ten workflows: positioning validation, competitive battlecards, pricing research, product messaging testing, voice of customer research, customer segmentation, GTM strategy validation, content marketing engine, sales enablement, and product launch research. Each workflow takes minutes from instruction to finished deliverable.

Related Topics

Product MarketingSynthetic Market ResearchUse CasesBest PracticesMessaging Positioning