sgit newsroom v0.1.29 · snapshot 2026-09-24

Reading room · nfrs.sgit.ai · llms-full

On this page

Reading room / nfrs.sgit.ai / llms-full.txt · section 7 of 12

02 — The memory thesis: these sites are agent memory, done better

The commissioning message contains the clearest statement of the network's purpose ever made:

"the idea of these sites is to provide good briefs for agents to learn about how we work and think — these sites are a more evolved and focused version of what is usually called LLM memory."

This document develops that into the page it deserves — and recommends it also appear on sgit.ai, because it explains all fourteen sites at once.


1. The claim

What products call "memory" today is an accumulation: embeddings of past conversations, retrieved by similarity, private to one vendor's silo, unversioned, uncurated, and invisible to the person it describes.

The *.sgit.ai network is the same function built as publishing:

Conventional LLM memoryThese sites
Contentaccumulated transcriptscurated briefs — written, reviewed, pruned
Retrievalsimilarity search, opaqueaddressable URLs + llms.txt + llms-full.txt
Versioningnonegit, version-prefixed filenames, dated pages
Truthfulnesswhatever was saidthe reality-document rule: "briefs are aspirations, not facts"
Portabilitylocked to one vendorany agent, any vendor, one HTTP GET
LicenceunclearCC BY 4.0, explicit, irrevocable
Inspectable by the humanrarelyit is a website — read your own memory
Shared across agentsnoyes — one memory, N agents, no sync

The one-line version for the front page: memory you can read, cite, version, license and hand to any agent — because it is a website.

2. The evidence that this was already the design

The estate has been building toward this explicitly, before the sentence was said:

3. What "more evolved" concretely means — the disciplines this site teaches

The memory thesis is an NFR story, which is why this page lives here: a memory is only better than a transcript if somebody maintains its non-functional requirements.

  1. Curation is the villagers' work applied to knowledge. Someone owns each site, prunes it, and keeps it true — the reality-doc discipline as memory hygiene.
  2. Versioning makes memory correctable — the thing a frozen vault and a transcript both cannot do.
  3. The do-not-publish tiers are memory's security model. Every pack in this series ships one; a memory without redaction discipline leaks.
  4. shipped/ pages are memory's calibration. A memory that cannot distinguish designed from built teaches agents to overclaim — the single most common failure this session found across the estate's stale artefacts.
  5. Deconfliction is memory's normalisation. One canonical copy, siblings link — the skills pack's diverging duplicate is what happens otherwise.

4. The honest limits


This document is released under the Creative Commons Attribution 4.0 International licence (CC BY 4.0).


== briefs/v0.33.62__nfrs-brief-pack__03__the-owned-disciplines.md