prompt engineering

This entire page is part of IBM's prompt engineering guide and is structured around prompt engineering techniques.

Prompt engineering is the process of structuring inputs, and it has emerged as a crucial technique for maximizing the utility and accuracy of these models" — Direct definition and articulation of prom

Instead of creating the perfect prompt, context engineering focuses on optimizing both the prompt and everything around the prompt." — Article explicitly positions context engineering as the evolution

clever prompts represented perhaps 0.1% of the total context modern AI systems process" — Article directly challenges prompt engineering as the primary driver of AI effectiveness in production. Contex

Context engineering combines prompt engineering, retrieval-augmented generation (RAG), and multi-agent techniques into one system, instead of using them separately." — Article explicitly names prompt

Over the past couple of years, building applications with large language models (LLMs) has shifted focus from prompt engineering to context engineering. In early LLM applications, users spent time cra

是一套面向 AI 行为工程的提示设计方法论，用最小的上下文开销，换取最可靠的行为改变" — agent-skills defines a prompt design methodology that minimizes context overhead while maximizing behavioral reliability — extends prompt engineering with

Explicitly mentioned as Chapter 3 of the curriculum; prompt engineering is foundational to context design

[DIRECT] "Putting "no eyebrows" in my CLAUDE.md" — Author uses CLAUDE.md configuration file to specify behavioral modifications (persona traits like eyebrows), demonstrating practical system prompt us

Layer around the model: defense is a system layer, not fine-tuning, modeled on software security." — Article explicitly describes defense as system-layer architecture, modeled on software security pri

you can get indirect prompt injection to ~0 on unseen attacks if you stack enough layers (model training + input probes + a classifier checking intent)" — Directly describes a proven multi-layered def

Reframes prompt engineering from 'writing longer prompts' to 'structuring information optimally for the model's representation space.'

The codebase is part of the prompt." — Frames codebase itself as essential prompt context, extending conventional prompt engineering to include code state as a first-class prompt component.

check every hypothesis against the recorded history in python before spending actions on it. additionally, every submitted action must carry an expected result, so a missed prediction cancels the rest

Every prompt must: Restate the problem exactly, Define what counts as a full solution, Explicitly ban weaker intermediate results, List traps & edge cases, Force independent adversarial agents to atta

As the models have gotten smarter, they need less direction, fewer constraints, and fewer examples." — Article documents Anthropic's strategic shift in prompt design: reducing system prompts by 80% as

As the models have gotten smarter, they need less direction, fewer constraints, and fewer examples." — Article demonstrates that modern LLMs require less explicit direction in system prompts; as model

"one entire C. elegans" is at the very top with almost 100% of images with a single mask at all developmental ages... "one whole C. elegans" is at THE VERY BOTTOM with only 4-25% of images having a si

The core problem is optimizing prompts to guide model behavior; demonstrates that small prompt variations cause large output differences

Reframes prompt engineering as context distribution problem rather than prompt verbosity problem

"像设计系统一样写 prompt"的方法论" — Article presents a systematic methodology for prompt engineering using design-system principles, introducing constraints-card approach with FONT/STYLE/MODE parameters

The practice of structuring prompts for better results is fundamental prompt engineering, but elevated here to a meta-level: the structure of how you introduce context matters more than the quality of

[direct] "When a better AI model gets released, the first thing you should try is to REMOVE instructions." — Article challenges conventional prompt engineering wisdom by advocating for FEWER instructi

The GCOB framework is a structured prompt engineering methodology

Author's 'clear stop points' are prompt-level engineering patterns that directly control context consumption

Article explicitly distinguishes context engineering (complete information environment) from prompt engineering (single-turn phrasing). Frames context engineering as superset addressing production req

Article explicitly identifies prompt engineering as the 2023 era and establishes context engineering as its successor, creating clear conceptual lineage

Markdown gives you headings, code blocks, bold/italic text, and organized lists. This structure helps AI (and us) quickly find sections without having to read the entire file top to bottom." — Article

Context Engineering is the new big thing for AI coding - not because the concept is new, but because we're finally reaching the point where everyone realizes that putting in the effort to provide exte

Prompts are great for one-off requests and human-in-the-loop interfaces, but terrible for defining the behaviors of systems" — Article directly challenges the applicability of prompts for system-level

Article treats prompt engineering (persona + instructions) as one component of broader context engineering discipline, positioning context engineering as the superset that includes data, state, retrie

Article explicitly identifies prompt engineering as critical skill for multi-agent design; context/role/persona defined via prompts determine agent behavior.

Article explicitly positions context engineering as distinct from and broader than prompt engineering - a key boundary definition. Context engineering encompasses the environment; prompt engineering i

Article explicitly frames context engineering as evolution/progression from prompt engineering, showing how the earlier concept remains valid but becomes secondary to context design.

Article explicitly positions context engineering as the evolution of prompt engineering, incorporating prompt strategies but expanding to holistic token management

[direct] "a leitwort in AI is any word or phrase you use that appears in the agents' thinking traces and guides its behavior" — Introduces leitwörter as a novel prompt engineering technique borrowed f

the questions they ask" — Article directly states that expert users get more value from Claude because of the quality of questions they ask, which is the core of prompt engineering skill.

Article explicitly frames context engineering as evolution beyond prompt engineering, positioning the latter as 'battle against complexity'

Jordan Crawford audited 810 prompts he typed into Claude Code over 17 days to understand how he interacts with the AI" — Article documents a real-world prompt engineering practice with quantified metr

Session explicitly covers 'automated prompt development' as a context engineering lever. Prompt engineering is a foundational CE technique.

Author explicitly identifies prompt engineering as the turning point from failure to success in 2024→2025

Setting a role in the system prompt anchors tone and persona. Anthropic's prompt-engineering guidance is especially clear that role + clear instruction structure beats clever wording." — Article expli

Article explicitly positions context engineering as a replacement/displacement of prompt engineering, arguing that prompt tuning alone fails at scale for stateful systems with tools and agents.

The article demonstrates specific prompt design (adversarial framing) as a context engineering technique

2025: Ask AI better" — Article explicitly identifies prompting/prompt engineering as the 2025 competitive advantage in AI interaction

Context Engineering, a formal discipline that transcends simple prompt design to encompass the systematic optimization of information payloads for LLMs" — Survey explicitly positions context engineeri

Episode explicitly distinguishes context engineering from prompt engineering, treating them as separate disciplines with different focuses and skills

writing prompts (prompt engineering is just a small piece of it)" — Directly positions prompt engineering as subset of broader context engineering discipline, adding novel dimension that context encom

[direct] "I really like the term 'context engineering' over prompt engineering. It describes the core skill better: the art of providing all the context for the task to be plausibly solvable by the LL

GEPA extends static prompt engineering into dynamic, feedback-driven prompt evolution. Shifts from 'craft good prompt' to 'system that improves prompts iteratively'.