0

Discovery Workflow

awesome-llm-appsreferences

Goal: Idea maturation, not speed.

Sign in to save downloads to your library and vote.

Preview

Discovery Workflow

Goal: Idea maturation, not speed.

Proceed in phases. Even if the user provides details, complete each phase through conversation. Do not infer or assume -- discuss and validate with user. Proceed one phase at a time.


Phase 1: Value Proposition

  1. Problem + User: What problem does this solve? For whom?
  2. Pain: How is it solved today? What's painful about that?
  3. Core actions: 1-3 focused actions (not a full app port)

Phase 2: Why LLM?

  1. Conversational win: Where does "just say it" beat clicking through menus?
  2. LLM adds: What does the LLM contribute? (intent parsing, generation, reasoning, summarization)
  3. What LLM lacks: Your data? Your APIs? Ability to take real actions?

Fail patterns (stop if any match):

  • Long-form or static content better suited for a website
  • Complex multi-step workflows that exceed widget display modes
  • Dashboards (use tables, lists, or short summaries instead)
  • Full app ports instead of focused atomic actions
  • No clear answer to "why inside an AI assistant vs standalone?"

If a fail pattern matches: explain the gap, suggest a different interface or narrower scope.


Phase 3: UI Overview

Describe the user journey through core actions:

  1. First view: What does the user see when the widget loads?
  2. Key interactions: What happens at each core action?
  3. End state: How does the experience conclude?

Phase 4: Product Context

Gather:

  • Existing products and APIs
  • Authentication method (OAuth, API key, none)
  • Rate limits and constraints
  • Data sources

Phase 5: Spec Summary

After phases 1-4 are discussed, assemble a brief spec:

# App Name

## Value Proposition

[Problem, user, pain point, core actions]

## Why LLM?

[Conversational win, what LLM adds, what it lacks]

## UI Overview

[First view, key interactions, end state]

## Product Context

[APIs, auth, constraints]

## Tools and Widgets

[List of tools and widgets with input/output shapes]

After the spec is ready, proceed to architecture.md for API design, then setup.md for project scaffolding.

Ingestion metadata

Source catalog
awesome-llm-apps
File path
generative_ui_agents/ai-mcp-app-builder/apps/mcp-use-server/.agent/skills/chatgpt-app-builder/references/discover.md
Last refreshed
10/24/2026, 3:00:14 AM (5h ago)
Refresh schedule
Daily · 03:00 UTC
Dedupe status
Unique · deduped by (source, url)