Claude Skill

product-context

Build or refresh a compact, evidence-backed product context before marketing, positioning, SEO, analytics, sales-enablement, or launch work. Use when an agent needs to understand the product, audience, proof, conversion goal, terminology, and claim boundaries without inventing co

LLM Mart · 0 points · 3 views 0 listing impressions 0 install-command copies
Download markoblogo-abvx-agent-skills-skills_product-context-37a7146.zip · 2 KB

Install

skills CLI npx skills add https://github.com/markoblogo/abvx-agent-skills/tree/main/skills/product-context
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install markoblogo-abvx-agent-skills@llmmart
Git git clone https://github.com/markoblogo/abvx-agent-skills.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole markoblogo/abvx-agent-skills collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Product Context

Create a small, durable context artifact that lets later product work start from evidence instead of generic positioning.

Use When

  • entering marketing, content, SEO, analytics, sales, or launch work in a repo;
  • the product, audience, conversion action, or proof boundary is unclear;
  • existing public copy and internal product behavior have drifted.

Workflow

  1. Read existing product docs, public copy, analytics contracts, and relevant implementation surfaces.
  2. Separate confirmed facts from assumptions, placeholders, and claims needing owner confirmation.
  3. Draft or update docs/product-context.md unless the repo already defines a better canonical path.
  4. Keep only: product and audience, jobs and conversion action, terminology, proof and known limits, approved voice, prohibited claims, source surfaces, and last-reviewed date.
  5. Ask for confirmation where positioning, ICP, pricing, competitor claims, or proof cannot be established from the repo.
  6. Link future work to the context; do not treat it as permission to publish, spend, contact people, or change product behavior.

Guardrails

  • Never invent customer quotes, metrics, market position, availability, pricing, partner status, or legal/compliance claims.
  • Keep personal data and private sales evidence out of a committed context file.
  • Treat a product context as a reviewable input, not a source of truth over code, current policy, or human direction.
  • Preserve domain-specific safety language such as financial, booking, or approval boundaries.

Output Shape

Use these headings where applicable: Product, Audience and jobs, Conversion action, Proof and limits, Language, Claim boundaries, Source surfaces, Maintenance.

Pair With

  • evidence-ledger-research for external facts;
  • social-publishing-gate before distribution;
  • bounded-growth-loop before recurring optimization;
  • loop-readiness-review before scheduling any loop.

Provenance

Adapted from the shared-context architecture in coreyhaines31/marketingskills, narrowed to evidence-backed, product-neutral context with explicit claim boundaries.

Files (abvx-agent-skills)
  • agents
    • openai.yaml 222 B
      Loading…
  • SKILL.md 2.5 KB
    ---
    name: product-context
    description: Build or refresh a compact, evidence-backed product context before marketing, positioning, SEO, analytics, sales-enablement, or launch work. Use when an agent needs to understand the product, audience, proof, conversion goal, terminology, and claim boundaries without inventing commercial facts.
    license: MIT
    metadata:
      abvx_status: experimental
      abvx_origin: adapted
    ---
    
    # Product Context
    
    Create a small, durable context artifact that lets later product work start from evidence instead of generic positioning.
    
    ## Use When
    
    - entering marketing, content, SEO, analytics, sales, or launch work in a repo;
    - the product, audience, conversion action, or proof boundary is unclear;
    - existing public copy and internal product behavior have drifted.
    
    ## Workflow
    
    1. Read existing product docs, public copy, analytics contracts, and relevant implementation surfaces.
    2. Separate confirmed facts from assumptions, placeholders, and claims needing owner confirmation.
    3. Draft or update `docs/product-context.md` unless the repo already defines a better canonical path.
    4. Keep only: product and audience, jobs and conversion action, terminology, proof and known limits, approved voice, prohibited claims, source surfaces, and last-reviewed date.
    5. Ask for confirmation where positioning, ICP, pricing, competitor claims, or proof cannot be established from the repo.
    6. Link future work to the context; do not treat it as permission to publish, spend, contact people, or change product behavior.
    
    ## Guardrails
    
    - Never invent customer quotes, metrics, market position, availability, pricing, partner status, or legal/compliance claims.
    - Keep personal data and private sales evidence out of a committed context file.
    - Treat a product context as a reviewable input, not a source of truth over code, current policy, or human direction.
    - Preserve domain-specific safety language such as financial, booking, or approval boundaries.
    
    ## Output Shape
    
    Use these headings where applicable: `Product`, `Audience and jobs`, `Conversion action`, `Proof and limits`, `Language`, `Claim boundaries`, `Source surfaces`, `Maintenance`.
    
    ## Pair With
    
    - `evidence-ledger-research` for external facts;
    - `social-publishing-gate` before distribution;
    - `bounded-growth-loop` before recurring optimization;
    - `loop-readiness-review` before scheduling any loop.
    
    ## Provenance
    
    Adapted from the shared-context architecture in `coreyhaines31/marketingskills`, narrowed to evidence-backed, product-neutral context with explicit claim boundaries.
    
  • SKILL_CARD.md 1.4 KB
    Loading…

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.