15:I[981,["658","static/chunks/658-a698693f56d0116a.js","183","static/chunks/183-6beeeb801c274aaa.js","474","static/chunks/app/%5Btype%5D/layout-631c4730b83c8223.js"],"Posts"] 16:T7c4,Hi HN,

I’m Akshay. I built AMP because I was tired of my AI agents having "amnesia" the moment I closed the terminal.

Like many of you, I use Claude/Cursor daily. RAG is great for searching documentation, but it’s terrible for continuity. It chunks text blindly, losing the narrative. When I asked my agent "Why did we decide to use FastAPI last week?", it would hallucinate or just give me generic pros/cons because the specific context of our decision was lost in a vector soup.

So I decided to build a proper *Hippocampus* for my agents.

*What is it?* AMP is a local-first memory server that sits between your agent and its LLM. It implements a "3-Layer Brain": 1. *STM*: A scratchpad for what we are doing right now. 2. *LTM*: Consolidated facts and insights (promoted from STM). 3. *Galaxy Graph*: A force-directed knowledge graph (d3.js) that physically links related concepts.

*The Tech Stack* * *Protocol*: Built natively for the *Model Context Protocol (MCP)*. If you use Claude Desktop or Cursor, it works out of the box. * *Backend*: Python, FastAPI, SQLite (no docker needed). * *Search*: Hybrid (Keyword + Vector) using FastEmbed. * *Viz*: A 60fps local dashboard to actually see your agent's brain growing.

*Benchmarks* I didn't want just another vector wrapper, so I focused heavily on retention. I benchmarked it against Mem0 (a popular alternative) on the LoCoMo dataset. * *AMP*: 81.6% Recall (Context-First) * *Mem0*: 21.7% Recall (Extraction-First)

It turns out that preserving the narrative (who/what/when) before summarizing is key to avoiding "I don't know" answers.

I’d love to hear your thoughts on the architecture. Does your agent need a hippocampus?

Repo:

First – For Inspiration and Recognition of Science and Technology

Show HN: Brandcaster AI – On-brand Content Creator and Posting Agent

The claims of close reading

Tinker: General Availability and Vision Input

Show HN: SigmaLifting – A protocol for powerlifting training data

Anonymous VPN: Tor and Mullvad VPN

New React vulns leak secrets, invite DoS attacks

Show HN: AMP – open-source memory server for AI agents (MCP, SQLite, D3.js)

AI superintelligence is a Silicon Valley fantasy, Ai2 researcher says

Laptop and Gaming Handheld Device Improvements Merged for Linux 6.19

Loop-Invariant Code Motion

3

Show HN: Full Signal – Get the signal of Twitter without doomscrolling

China is worried about AI job losses

Openfire 5.0.3 Released – Open-Source – Java XMPP/Jabber Server

Brown University shooting: At least 2 killed

Show HN: Binaural Beat Player

I miss the old Internet of 10-20 years ago

TeXPen

Copywriters reveal how AI has decimated their industry

Show HN: Vibe coded 3D Scene Animation Studio

The Core Misconception That Is Driving American AI Policy

From Zero to QED: An informal introduction to formality with Lean 4

Closures as Win32 Window Procedures

Smooth Scrolling on the Sega Master System

Show HN: I built a service to fix NPM's email flood problem

What Happens When Pathogens Get Smaller Than Viruses? [video]

Yale Shooting Problem

Thirty Slices/Twenty-Four Days: How Christmas Was Saved by Abandoning Estimation

What is the most effective way to learn programming?

Show HN: I built a small tool to track receipts and warranty dates