Skip to content

phuryn/product-engineering / app-stack/skills/embedded-agents-messages-api

embedded-agents-messages-api

Verified

Building, adding or wiring up an agent or an agent chat inside a Next.js app: embed Claude agents on the Messages API, the default way.

examples/knowledge-agent/knowledge/product-vision.md

View on GitHub
# Product Vision

## What it is

A product vision is a **vivid, durable description of the future state the product is trying to create**. It is the answer to: *If we succeed completely over the next 5–10 years, what does the world look like, and what role does this product play in it?*

A vision is qualitative, opinionated, and long-lived. It does not change with each quarterly plan. It is the fixed star that strategy navigates by.

## Vision vs. mission vs. strategy

These three are routinely confused.

- **Mission** — why the company exists. The reason for being, beyond profit. ("Organize the world's information.") Mission is timeless.
- **Vision** — what the world looks like when the mission is being fulfilled through *this* product. Vision is concrete enough to picture; mission is abstract.
- **Strategy** — the time-bound bets being made now to move toward the vision. (See `product-strategy.md`.)

Order of operations: mission → vision → strategy → roadmap. A vision that is just a paraphrased mission is too abstract. A vision that is just a longer roadmap is too short-term.

## Characteristics of a strong product vision

- **Pictorial.** A reader should be able to *see* the future you're describing. Use scenes, not adjectives. "A small business owner in 2030 closes the books in 20 minutes from their phone" beats "We will be the leading SMB accounting platform."
- **Opinionated about who matters.** A vision implies a specific customer whose life is changed. Vague "everyone" visions are forgettable.
- **Durable.** It should still be true 5 years from now without rewriting. If a competitor's launch invalidates it, the vision was actually a strategy.
- **Inspirational without being unfalsifiable.** "Make the world a better place" can never be wrong, and therefore guides nothing. A good vision could, in principle, be the *wrong* future to aim at — that's what makes it a real commitment.
- **Short enough to remember.** If the team can't recite it from memory, it can't guide decisions.

## What a vision is for

A product vision is a **decision-making tool**, not a marketing artifact. Its purpose:

1. **Recruiting.** Senior people choose where to work based on the future being built, not the current backlog.
2. **Prioritization.** When two roadmap options are roughly equal, the one that moves closer to the vision wins.
3. **Resilience.** During hard quarters, the vision is what keeps the team committed to the long bet instead of optimizing the short one.
4. **Alignment across functions.** Engineering, design, sales, and marketing make better-coordinated choices when they share a picture of the destination.

## Common failure modes

- **Vision is a brochure.** Written for investors, not for the team. Polished, lifeless, never referenced after the all-hands.
- **Vision is a roadmap with longer dates.** "In 2027 we will have shipped X, Y, and Z." This is just a plan, not a picture.
- **Vision is everything to everyone.** Tries to serve every customer, win every market, do every thing. Implies no trade-offs, guides no decisions.
- **Vision is hidden.** Lives in a doc nobody opens. Effective visions are repeated until they become slightly embarrassing — that's roughly the right amount.
- **Confusing vision and ambition.** "Be the biggest" is an ambition. "Knowledge workers spend zero time on filing or finding documents" is a vision — it describes a *world*, not a *position*.

## How to test a draft vision

Ask three questions:

1. **Could a competitor write the same vision?** If yes, it's too generic.
2. **Does it imply specific decisions?** Walk through 3 recent roadmap trade-offs. Does the vision actually help pick between them? If not, it's too vague.
3. **Will it still be true in 5 years?** If a foreseeable technology shift would invalidate it, it's a strategy in disguise.

A vision that survives all three is doing work.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall