product-marketing-context

Overview of product-marketing-context skill

What product-marketing-context does

The product-marketing-context skill helps you create and maintain a shared product marketing source file at .agents/product-marketing-context.md. Its job is simple but valuable: capture your product overview, audience, positioning, objections, proof points, and messaging basics once so other marketing workflows can reuse them instead of asking the same foundational questions every time.

Who should use this skill

This is best for founders, product marketers, growth leads, and agencies who need repeatable AI-assisted marketing work across a repo or client project. It is especially useful if you want better consistency across messaging tasks like positioning, ICP work, landing page drafts, campaign planning, or sales-enablement content.

The real job to be done

Most teams do not need “more copy.” They need a stable context layer the model can reference before it writes anything. The product-marketing-context skill is designed for that setup step. It creates a reusable marketing context document first, then other skills can build on it.

Why this is better than a generic prompt

A normal prompt can ask for positioning, but it usually produces a one-off answer. product-marketing-context is structured around:

• checking for an existing context file first
• updating only the sections that changed
• offering an auto-draft from the codebase or a guided start-from-scratch flow
• storing the result in a predictable file path other skills can reference

That makes it more operational than a generic “help me with messaging” prompt.

Best-fit and misfit cases

Best fit:

• new products with no messaging foundation yet
• repos where multiple AI marketing skills will be used
• teams updating positioning after a launch, pricing change, or audience shift

Weak fit:

• one-off copy tasks where you do not want persistent context
• products with no accessible repo or product materials to inspect
• teams expecting deep market research from the skill alone

How to Use product-marketing-context skill

Install context and where the file lives

If your environment supports Skills installation, add the parent repository and use the product-marketing-context skill from there. The output file this skill is built around is:

/.agents/product-marketing-context.md

The skill also checks .claude/product-marketing-context.md for older setups and can help migrate that content forward.

Start by checking for an existing context file

Before creating anything new, the intended workflow is to look for:

• .agents/product-marketing-context.md
• .claude/product-marketing-context.md

If a context file already exists, the right move is not “rewrite everything.” Read it, summarize what is already captured, then update only the sections affected by the user's changes.

Choose the right setup path

The repository shows two practical entry paths:

1. Auto-draft from codebase
   Recommended when the repo has usable signals such as README.md, landing page copy, docs, package.json, or product copy in-app.

2. Start from scratch
   Better when the repo is thin, private materials matter more than code, or the product is still being defined.

This is one of the most important adoption details in the product-marketing-context guide: you get better results if you choose the path that matches your evidence quality.

What input product-marketing-context needs

To get a useful output, provide at least:

• what the product is
• who it is for
• the main problem it solves
• how it differs from alternatives
• any current proof points or constraints
• whether this is a first draft or an update

Useful extra input:

• pricing tier changes
• new personas
• win/loss patterns
• customer language from calls or reviews
• brand voice preferences
• sales objections
• launch or quarter goals

The sections this skill is trying to build

Based on the repository evals and workflow, expect the product-marketing-context usage flow to cover sections like:

• Product Overview
• Target Audience
• Personas
• Problems You Solve
• Competitive Landscape
• Differentiation
• Objections
• Switching Dynamics
• Customer Language
• Brand Voice
• Proof Points
• Goals

That structure matters because it creates a durable context file, not just a positioning paragraph.

Turn a rough request into a strong prompt

Weak:

Help me with product marketing.

Better:

Use the product-marketing-context skill. Check whether .agents/product-marketing-context.md already exists. If not, auto-draft a V1 from the repo and website copy. Capture product overview, ICP, personas, key pain points, differentiation, objections, proof points, customer language, and brand voice. Flag assumptions clearly so I can review them.

Best for updates:

Use the product-marketing-context skill to update our existing context. We added an enterprise tier and now sell to VP of Engineering in addition to Product Managers. Read .agents/product-marketing-context.md, tell me which sections need revision, then update only those sections and preserve everything still valid.

A practical workflow that reduces guesswork

A good working sequence is:

1. Check for an existing context file.
2. Decide auto-draft vs. guided interview.
3. Build or update the context file section by section.
4. Mark assumptions and missing evidence.
5. Review with the user for corrections.
6. Save to .agents/product-marketing-context.md.
7. Reuse that file in later marketing tasks.

This is the main reason to do a product-marketing-context install early in a project: it lowers repeated setup work later.

Read these repository files first

For fast evaluation, start with:

• skills/product-marketing-context/SKILL.md
• skills/product-marketing-context/evals/evals.json

SKILL.md explains the operational workflow. evals/evals.json is useful because it shows what “correct usage” looks like in realistic prompts, including updates after product changes.

What materially improves output quality

The best outputs come from evidence, not aspiration. If you want a stronger context doc, point the skill toward:

• homepage or landing page copy
• product docs and onboarding text
• demo transcript notes
• pricing and plan descriptions
• customer interview quotes
• competitor comparisons you actually use
• sales-call objections
• existing brand guidelines

Without those, the draft may still be usable, but it will contain more generic assumptions.

What to expect from the first draft

A V1 created with product-marketing-context should be treated as a working document, not final truth. The most useful draft:

• separates confirmed facts from inferred positioning
• exposes gaps in proof points or audience clarity
• gives enough structure for downstream AI marketing tasks to stay consistent

If you want polished strategic messaging, expect one review pass after the initial file is created.

product-marketing-context skill FAQ

Is product-marketing-context good for beginners?

Yes. It is beginner-friendly because the workflow is explicit: check for an existing file, choose auto-draft or guided setup, then fill specific sections. You do not need to know formal PMM frameworks to get a usable first version.

When should I use product-marketing-context instead of a normal prompt?

Use it when you want reusable context across multiple tasks. If your goal is one landing page headline, a normal prompt may be enough. If your goal is consistent messaging across many outputs, product-marketing-context for Product Marketing is the better foundation.

Can it update an existing context instead of recreating it?

Yes. That is a core behavior. The skill is designed to read an existing .agents/product-marketing-context.md, summarize what is already there, and revise only the sections affected by new information.

Does it do market research for me?

Not really. It can infer a lot from the repo and product materials, but it is not a substitute for customer research. If your evidence is weak, the output will be structurally helpful but strategically thinner.

What if my repo has very little marketing copy?

Use the start-from-scratch path and answer the questions conversationally. This is often better than forcing an auto-draft from sparse codebases that reveal little about customers, objections, or buying context.

Is product-marketing-context only useful inside this repo ecosystem?

It is most valuable when paired with other marketing skills that reference .agents/product-marketing-context.md, but the file itself is portable. You can still use it as your canonical PMM brief even outside the full skill ecosystem.

When is this skill a poor fit?

Skip it if:

• you only need a single isolated copy output
• your team will not maintain a shared context file
• your product is still too undefined to answer basic audience and problem questions

How to Improve product-marketing-context skill

Give the skill evidence, not just goals

The biggest upgrade to product-marketing-context skill quality is better source material. “We want stronger positioning” is not enough. Provide raw inputs such as:

• homepage text
• customer call notes
• top objections from sales
• why customers switch
• feature-to-benefit explanations
• proof points with numbers

That lets the model anchor claims instead of inventing them.

Be explicit about what changed

For updates, do not ask for a broad refresh if only two things changed. Say exactly what is new:

• new pricing tier
• expanded ICP
• renamed category
• stronger enterprise focus
• new competitor pressure

This helps the skill update the right sections without accidentally weakening still-correct messaging.

Ask for assumptions to be labeled

One common failure mode is hidden guesswork. Improve output quality by asking the model to mark:

• confirmed facts
• likely inferences
• missing inputs
• questions for review

That makes the first draft far easier to approve and refine.

Use section-by-section review for better final output

Do not review the entire document only at the end. The highest-value sections to validate early are:

• Target Audience
• Problems You Solve
• Differentiation
• Objections
• Proof Points

If those are wrong, later messaging built on top of the file will also drift.

Strengthen customer language capture

A frequent weak spot is generic wording. To improve the product-marketing-context guide in practice, feed it actual phrases customers use in:

• interviews
• support tickets
• review sites
• sales calls
• onboarding feedback

Real customer language sharply improves downstream copy relevance.

Watch for these common failure modes

Typical issues include:

• feature lists disguised as positioning
• overly broad ICPs
• weak differentiation against vague “competitors”
• aspirational brand voice with no examples
• proof points that are claims, not evidence

If you see these, revise the source inputs before requesting a rewrite.

Iterate with targeted follow-up prompts

After the first draft, use focused prompts like:

• “Narrow the ICP to the most likely first buyer.”
• “Rewrite objections based on enterprise procurement concerns.”
• “Separate direct competitors from status-quo alternatives.”
• “Turn proof points into evidence-backed bullets only.”

Targeted iteration improves the document much faster than asking for a full rewrite.

Maintain product-marketing-context as a living file

The best long-term use of product-marketing-context is ongoing maintenance, not one-time setup. Revisit the file when your product, audience, pricing, category, or sales motion changes. That keeps every downstream marketing task aligned with current reality instead of stale assumptions.