we view context engineering as the natural progression of prompt engineering. Prompt engineering refers to methods for writing and organizing LLM instructions ... Context engineering refers to the set of strategies for curating and maintaining the optimal set of tokens (information) during LLM inference, including all the other information that may land there outside of the prompts.
The process of managing this information flow is called context engineering. Context engineering includes strategies such as isolating context for a specific agent, persisting information across multiple steps, or compressing large amounts of data to improve efficiency.
Make MCP stateless: remove the `initialize`/`notifications/initialized` handshake. Every request now carries its protocol version and client capabilities in `_meta`... 1. Deprecate the Roots, Sampling, and Logging features (SEP-2577).
For models designed with explicit reasoning capabilities, CoT prompting often results in only marginal, if any, gains in answer accuracy. However, it significantly increases the time and tokens needed to generate a response. ... o3-mini (2.9%) and o4-mini (3.1%); Performance decrease for Gemini Flash 2.5 (-3.3%).
On high-depth P-complete tasks (GSM8K, MATH), CoT gives a +54 to +68 pp recovery gap across all models. On shallow TC^0 tasks (MMLU, ARC), CoT is structurally redundant (Delta in [0.0, +4.6] pp).
For high-performance LLMs, simple prompts often outperform more complex ones while reducing cost. In contrast, commonly used prompting styles in natural language processing, such as step-by-step reasoning, or the use of reasoning models often lead to lower accuracy.
We study three optimizers from the DSPy framework -- COPRO, MiPROv2, and SIMBA ... MiPROv2 yielding the most stable gains for CoT, and SIMBA providing the largest benefits for ReAct agents, particularly at larger model scales. ... CoT remains effective for tabular fact checking, especially with smaller models.
these BetterTogether strategies optimizing the weights and prompts of a pipeline together outperform directly optimizing weights alone and prompts alone by up to 60% and 6%, respectively
Store production prompts in your application code instead of creating reusable prompt objects. ... OpenAI is deprecating reusable prompt objects in the API. Prompt creation will be de-emphasized beginning June 3, 2026, and `v1/prompts` is scheduled to shut down on November 30, 2026.
Wrap examples in `<example>` tags (multiple examples in `<examples>` tags) so Claude can distinguish them from instructions. ... XML tags help Claude parse complex prompts unambiguously, especially when your prompt mixes instructions, context, examples, and variable inputs.
Set `reasoning.effort` to `low`, `medium` (default), `high`, `xhigh`, or `max`. ... When reasoning effort is not `none`, remove `temperature`, `top_p`, and `top_logprobs`.