Field Report — not CI-gatedJuly 20, 2026

Measuring AI agent memory across a major refactor

A ~9,300-node TypeScript SaaS platform was rebuilt with NeuralMind watching. The memory didn't go stale — it grew into the new architecture. Here are the numbers, where each one came from, and how to run the same measurement on your repo.

Darren Frost (dfrostar)Full walkthrough on GitHub →

Avg token reduction

48.8×

~1,033 tokens/query vs 50K+ naive

Personal synapse edges

36 → 135

the learning layer tracked the new code

Full rebuild

326 s

then back to ~30 s incremental

The setting

A private, mid-size TypeScript SaaS platform (~9,300 indexed nodes), maintained by NeuralMind's own maintainer — see the honesty notes below. A major internal rebuild ("Phase 3 → Phase 4") landed a new backend module, a shared business-logic layer, an admin UI, and end-to-end specs.

NeuralMind ran throughout with lifecycle hooks installed, so the synapse layer observed the work as it happened. The repo is private, so the numbers are anonymized; every one of them came from the shipped CLI.

The numbers

MetricBefore (Phase 3)After (Phase 4)Change
Total nodes9,1909,293+103
Communities810new resolution — no comparable before
Personal synapse edges36135+275%
Shared synapse edges10,07910,183+104
Shared edge weight2,774.732,924.66+5.4%
Wake-up tokens455
Avg query tokens1,033
Avg token reduction48.8×vs loading files naively
Full --force rebuild326 s (~5.4 min)incremental rebuilds ~30 s after

What the numbers say

48.8× token reduction. After the rebuild, an average code question cost ~1,033 tokens of context instead of the 50K+ a naive "load the relevant files" approach would spend. That is one repo's measured ratio from neuralmind benchmark ., consistent with the 40–70× real-repo range — not a universal guarantee.

Personal synapse edges tripled (36 → 135). The Hebbian synapse layer learned co-activations across the new code as it was being written and queried. The memory didn't go stale through the rebuild; it grew into the new architecture.

Shared edge weight +5.4%. The static graph's cross-links got denser as the new shared business-logic layer connected previously separate areas — densification, not just growth: edge weight rose 5.4% on only 1.1% more nodes.

Rebuild cost stayed flat. One full --force rebuild at 326 s to pick up the new structure, then incremental rebuilds back to ~30 s. Re-indexing is not a tax you pay per edit.

Where each number comes from

Honesty first: the before/after table is hand-assembled from the output of several commands — there is no single neuralmind compare-phases command.

MetricCommand
Total nodes, communitiesneuralmind stats . --json
Personal/shared edges, edge weightneuralmind stats . — the Memory namespaces block
Wake-up / query tokens, reductionneuralmind benchmark .
Rebuild timetimed neuralmind build . --force
Synapses fired per queryneuralmind metrics .
Real logged spendneuralmind savings .

Run the same measurement on your refactor

1. Snapshot before you start

neuralmind stats . --json > .neuralmind-before.json

neuralmind benchmark . # baseline reduction number

2. Refactor with the memory watching

The synapse layer only learns from what it observes — install hooks before the work starts.

neuralmind install-hooks .

neuralmind watch & # optional: always-on learning from edits

3. Rebuild and snapshot after

time neuralmind build . --force

neuralmind stats . --json > .neuralmind-after.json

4. Re-benchmark and price it

neuralmind benchmark . # reduction on the post-refactor graph

neuralmind savings . # what your logged queries cost vs naive

Diff the two JSON files. The interesting deltas: node count (did the graph track the new code?), personal edges (did the memory learn the new co-activations?), shared edge weight (did the graph get denser or just bigger?). The full step-by-step walkthrough lives in the docs.

Honesty notes

One repo, one developer, measured by the maintainer on a private codebase. It is an existence proof and a recipe, not an independent benchmark. You cannot re-run it on the source repo — but you can run the identical method on yours, which is the point.

48.8× is a reduction in retrieval input tokens, not your total LLM bill. For typical end-to-end agent sessions the realistic total-cost saving is smaller — see the honest assessment.

The CI-gated, reproducible numbers live in Benchmarks & Results and on the benchmarks section of this site; this page is deliberately labeled a field report.