memoception

Boilerplates.

Every example, ready to read, download, and share — self-contained, zero-dependency, Node ≥ 18. Each one runs against the developer API with an API key from the app (settings → Developers). Click explore code to read any file, download it, or copy link to share a specific one.

Boilerplates

Custom-format store
store · rules · organize · 8 files

The pattern this boilerplate teaches:

Research agent
research · recall · consolidate · 3 files

A deterministic research loop that turns a subject + a list of questions

Multi-project knowledge
isolation · aggregation · 4 files

One project per subject, each a private, isolated memory partition, with a

Architecture (Canada)
precise plans · conventions · 7 files

For architects and BIM/CAD tooling working to Canadian conventions. Two

Unreal Engine optimization
headless · profiling · recall · 5 files

A headless UE optimization pipeline with a persistent brain. The engine

Godot optimization
headless · conventions · 5 files

The Godot sibling of the Unreal boilerplate: a headless Godot pipeline with a

Test suites & scripts

API harness
contract test · 1 file

// Memoception API v1 testing harness — exercises every endpoint against a live

Custom store/recall/organize suite
rich data + assertions · 1 file

// Custom store / recall / organize — comprehensive suite with RICH real data

Cross-endpoint suite
every endpoint · 1 file

// Cross-endpoint contract suite — exercises EVERY v1 endpoint the custom suite

New-endpoints suite
get/patch · query · batch · 1 file

// Suite for the six newest endpoints: GET/PATCH /memory/:id, GET /content/:id,

Web-file upload suite
every format, from the web · 2 files

// Upload real files found on the WEB across many formats, then verify the whole

Research script
aggregate a subject · 1 file

// Research boilerplate — aggregate online knowledge about a subject into a

Set your key once — export MEMOCEPTION_KEY=engli_agent_… — then run any script with node <file>. Full reference on the API docs page.