The articulation of intent to a machine, transforming a vague need into a request precise enough for a bot to act on meaningfully.
Intent
explicit and implicit…
Input mechanisms
Structured prompt
Reference materials with @mentions
Using @mentions to reference documents, images, code, and other materials provides context for better LLM responses.
Prompt guidance and scaffolding
Bridging the gap between fuzzy idea inside a human’s head and what makes an effective prompt for LLMs.
Quality feedback
Editing assistance
Templates
…
Suggestions
…
Prompt evaluation
…
Reflective prompting
Real-time, context-aware prompt completions derived dynamically from users’ immediate and historical interactions.
To-do
Precursors to capture: session, thread, and the other conversational containers a prompt sits within.
Resources & references
Related patterns
Precedes
- Bot — the prompt articulates the request the bot acts on
- Generated content — the input that leads to generated content.
- AI tuning — lets the actor refine how the model gets prompted
Enables
- Conversation — the prompt is the turn-level move a conversation with an AI is built from
Enacts
- Agency — the actor directs the system in their own words; initiative stays with whoever frames the request
Complements
- Searching — modern search inputs are evolving into prompt interfaces, accepting instructions alongside keywords
Preceded by
- Suggestion — can provide an initial prompt