With the latest GPT-5 I think it has done a great job at solving the needle in a haystack problem and finding the relevant files to change to build out my feature/solve my bug. Although, I still feel that it lacks some basic context around the codebase that really improves the quality of the response. Currently, the way agentic development works is that we do a semantic search using RAG (dense search) over our codebase and find the most relevant code or grep (sparse search) to solve our given problem/feature request.

I think that's great. But I also think that it gives room for improvement on how we think of context. Most time documentation is hidden in some architectural design review in a tool like notion, confluence, etc. Those are great for human retrieval but even then it is often time forgotten when we implement the code functionality. Another key issue is that as the code evolves, our documentation becomes stale.

We need a tool that follows the agentic approach we are starting to see where we have ever-evolving documentation, or memories, that our agents could utilize without another needle in a haystack problem.

For the past few weeks I have been building an open source MCP server that allows for the creation of "notes" that are specifically anchored to files that AI agents could retrieve, create, summarize, search, and ultimately clean up.

This has solved a lot of issues for me.

You get the correct context of why AI Agents did certain things, and gotchas that might have occurred not usually documented or commented on a regular basis.

It just works out-of-the-box without a crazy amount of lift initially.

It improves as your code evolves.

It is completely local as part of your github repository. No complicated vector databases. Just file anchors on files.

I would love to hear your thoughts if I am approaching the problem completely wrong, or have advice on how to improve the system.

You can find the project at https://github.com/a24z-ai/a24z-Memory.

Tony Blair Attends White House Meeting with Trump on Postwar Gaza

Engineer as Value Appraiser

Stop killing games: Demands for game ownership must also include workers' rights

Dead arms test importance of clenched fists (2015)

The Rise of Front-Loaded Vesting

The Cybersecurity Psychology Framework: A Pre-Cognitive Vulnerability Assessment

DraftKings said it acted properly in voiding Iowa man's $14.2M payout

Ask HN: What's your current "ecosystem"/dev stack/tools that you use to build?

What's new in Excel August 2025

Anthropic – Detecting and countering misuse of AI: August 2025

We're doing context engineering wrong

A retrospective about blogging for a decade

You're doing context engineering wrong

Asahi Linux Lead Developer Steps Down

Trying Out Claude for Chrome Research Preview

Flying Cameras Are Snapping Photos of Houses for Insurers

Lord of the Io_uring

Io_uring by Example

Layoff Tracker

Io_uring basics: Writing a file to disk

Observing the Earnings Gap Through Marital Status, Race and Gender (2019)

Atlassian's Trello redesign may be 'worst in tech history' say frustrated users

4chan and Kiwifarms Sue Ofcom over Attempt to Enforce Online Safety Act in US

Vibe Coding from My Smartphone

Show HN: DeepShot – an open-source NBA predictor with ML, EWMA, and live UI

The Electric Slide

Measuring founder's presence's contribution to stock price

Steiger: OCI-native builds for Docker, Bazel, and Nix with direct registry push

SOPS is an editor of encrypted files

4chan and Kiwi Farms file joint lawsuit against the UK