agentleFS
jiayuewangjavyjiayuewangjavy/product-designer-skills-os/skills/product-context/SKILL.md

product-context

Skill4 starsChanged 20 days agoMITView on GitHub

Create or update a shared, evidence-aware product context for product design and independent product-building work. Use when starting an initiative, consolidating scattered product inputs, or preparing reliable context for discovery, conceptual modeling, prototyping, evaluation, implementation, or marketing. Do not use as a substitute for missing user research or to silently turn assumptions into decisions.

8 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create the smallest durable context that helps downstream skills make consistent product decisions without repeatedly asking for the same …

    line 8
  2. Distinctive:

    Do not create a parallel source of truth when a current canonical document can be updated.

    line 14
  3. Distinctive:

    Use references/TEMPLATE.md when creating a new context file.

    line 31
  4. Distinctive:

    Use concrete information already provided.

    line 35
  5. Distinctive:

    Ask only when a missing choice materially changes the product.

    line 36
  6. Distinctive:

    When the user confirms or rejects something, update the decision record and retain the rejected option plus the reason when it can prevent …

    line 37
  7. Distinctive:

    Keep product context distinct from task instructions.

    line 38
  8. Distinctive:

    every consequential statement is labeled by epistemic status.

    line 45

“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Product Context

Create the smallest durable context that helps downstream skills make consistent product decisions without repeatedly asking for the same information.

Locate before creating

Inspect the user's supplied artifacts and the project's existing product brief, research, strategy, design-system, analytics, decision-log, and marketing-context files. Reuse the project's established location and terminology when one exists. Otherwise propose or create .agents/product-context.md when the task authorizes file changes.

Do not create a parallel source of truth when a current canonical document can be updated. Do not overwrite unrelated content or restructure an existing document beyond what the request needs.

Build the context

  1. Inventory every source actually inspected. Record its path or URL, date when known, and what it contributes.
  2. Extract only decision-relevant information into these classes:
    • Evidence: directly supported by research, behavior, analytics, policy, or an authoritative product source.
    • Inference: a reasoned interpretation of evidence.
    • Assumption: important but not yet verified.
    • Decision: explicitly chosen, with owner or source when known.
    • Preference: a user, brand, or team preference that is not a universal product principle.
  3. Describe the users and buyer separately when they differ. Include the critical job, current alternative, pain, desired outcome, and adoption constraints.
  4. Record the product promise, scope, non-goals, business model or value exchange, and meaningful technical, legal, accessibility, brand, and operational constraints.
  5. Capture unresolved questions and rank them by how much a different answer would change the product.
  6. Identify stale, contradictory, or weakly sourced context. Never silently resolve conflicts.
  7. End with the next decision and the minimum evidence needed to make it.

Use references/TEMPLATE.md when creating a new context file. Preserve an existing useful structure when updating one.

Interaction rules

  • Use concrete information already provided; do not ask the user to repeat it.
  • Ask only when a missing choice materially changes the product. Otherwise record it as an assumption or unresolved question and continue.
  • When the user confirms or rejects something, update the decision record and retain the rejected option plus the reason when it can prevent repeated debate.
  • Keep product context distinct from task instructions. This file records product truth and uncertainty; it does not prescribe how every downstream skill must work.
  • External research may add evidence, but popularity, competitor behavior, or a framework is not proof of the user's customer need.

Completion check

The context is ready when:

  • every consequential statement is labeled by epistemic status;
  • evidence has provenance;
  • user, buyer, problem, outcome, constraints, scope, and unknowns are findable;
  • contradictions are visible;
  • downstream work can name the next decision without inventing missing facts.

Return the saved path, the highest-risk unresolved assumption, and the recommended next skill or validation step.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.