Drew Breunig writes about applied AI, including context engineering, prompts, DSPy, and how language models are used in practice.
Simon Willison credits him with fantastic writing on context engineering, and commenters cite his posts on MCP, prompts, and LLM-as-judge.
100% yes! There've been some other writers who've been doing parallel work around that in the last couple weeks.
https://www.dbreunig.com/2025/06/10/let-the-model-write-the-prompt.html is an example.
You can see the hands on results in this hugging face branch I was messing around in:
here is where I tell the LLM to generate prompts for me based on research so far
https://github.com/AlexChesser/transformers/blob/personal/vibe-engineering-research/.devcontext/architecture/04-additonal-research-prompts.md#create-prompts-folder-and-per-recommendation-prompts
here is the prompts that produced:
https://github.com/AlexChesser/transformers/tree/personal/vibe-engineering-research/.devcontext/architecture/prompts
and here is the result of those prompts:
https://github.com/AlexChesser/transformers/tree/personal/vibe-engineering-research/.devcontext/architecture/onboarding. (also look at the diagram folders etc..)
So, this is cool and a per-user memory is obviously relevant for effective LLM use. And major props for the temporal focus.
However, keeping a tight, constrained context turns out to actually be pretty important for correct LLM results (https://www.dbreunig.com/2025/06/22/how-contexts-fail-and-how-to-fix-them.html).
Do you have a take on how we reconcile the tension between these objectives? How to make sure the model has access to relevant info, while explicitly excluding irrelevant or confounding factors from the context?
Came here to say this: people present MCP’s verbosity as all the context the LLM needs. But almost always, this isn’t the case.
I wrote recently, “ Connecting your model to random MCPs and then giving it a task is like giving someone a drill and teaching them how it works, then asking them to fix your sink. Is the drill relevant in this scenario? If it’s not, why was it given to me? It’s a classic case of context confusion.”
https://www.dbreunig.com/2025/07/30/how-kimi-was-post-trained-for-tool-use.html
indeed - https://www.dbreunig.com/2026/01/08/a-software-library-with-no-code.html appears to be exactly that - the idea that the only leverage you have for fixing bugs is updating prompts (and, to be fair, test cases, which you should be doing for every bug anyway) is kind of upsetting as someone who thinks software can actually work :-)
(via simonw, didn't see it already on HN)
Based on the recent leaks, their system prompt explicitly nudges the model not to do anything outside of what was asked. That could very well explain why it’s not fixing preexisting broken tests.
“Don't add features, refactor code, or make "improvements" beyond what was asked.”
https://www.dbreunig.com/2026/04/04/how-claude-code-builds-a-system-prompt.html
I would classify AnyTool as a context engineering trick. It's using GPT-4 function calls (what we would call tool calls today) to find the best tools for the current job based on a 3-level hierarchy search.
Drew calls that one "Tool Loadout" https://www.dbreunig.com/2025/06/26/how-to-fix-your-context.html#tool-loadout