Prompt
In one sentence
The instructions and reference materials provided to an AI, specifying what it should do, what information to use, and what format to follow.
The more specific the prompt, the more consistent the output. A good prompt usually clarifies four elements: context (the scenario), task (what needs to be done), reference materials (the data to rely on), and requirements (length, format, tone, and constraints).
When you repeatedly write prompts for the same type of task, you can standardize them into a structured instruction set for an AI agent to execute. That is a SKILL.
Example
"Based on the event plan below, write a formal National Day holiday notice under 300 words, ending with the on-duty phone number" is an example of a prompt.
Related terms
- System Prompt
- Pre-set instructions configured before a conversation begins that define the AI's role, rules, and response style. Users typically do not see them.
- SKILL
- A standardized operating procedure written for an AI agent that defines how to run a specific task, the sequence of steps, which materials to use, and the final delivery format.
- Context window
- The maximum total amount of text an LLM can process at once, including prompts, reference materials, chat history, and its own generated responses.
- Large Language Model (LLM)
- AI models trained on massive volumes of text that can understand and generate text. Doubao, DeepSeek, ChatGPT, and Claude are all powered by large language models.
Further reading
AI News
Why Code-Driven AI Agents Beat Prompt Engineering for Content
LangChain slashed content production costs from $8 to $0.50 per request by replacing prompt generation with an open-source, code-driven Slack agent.
Oct 28, 2026 · 5 min read
AI News
OpenAI Adds 12-Hour Prompt Caching: Stop Truncating Agent Context
OpenAI's 12-hour prompt caching removes the need to cut context to save costs. Anchor business rules in persistent prefixes for lower costs and stable runs.
Oct 21, 2026 · 4 min read
AI News
Universal AI Agents Are Here. Don't Hand Everything to a Prompt.
Unified AI entry points sound tempting, but without strict delegation boundaries, multi-system agents risk costly errors and blurred accountability.
Oct 9, 2026 · 4 min read
