Public REST API

A grammar for structured decisions.

The LAKA Volumetric Grammar treats any system as a sentence you can navigate: five change levels × ten internal variables × fourteen meta-variables — 700 base coordinates, each one a specific question about the system in front of you. This API serves all of it as JSON.

700 coordinates 6 operating modes 17 operators 50-cell grid No API key

Read the reference Open the explorer

Quickstart

One call, no key, no signup.

Every GET endpoint is public. Resolve a coordinate to see the shape of the data.

# Structural Change x Feedback x Acceleration
curl https://api.digitalstemcell.bowtiekreative.com/v1/coordinates/LAKA-C3-I07-M08

{
  "id": "LAKA-C3-I07-M08",
  "change_level":      { "code": "C3", "label": "Structural Change", ... },
  "internal_variable": { "code": "I07", "label": "Feedback", ... },
  "meta_variable":     { "code": "M08", "label": "Acceleration", ... },
  "grid_intent": "Multiple adaptive loops across the system",
  "prompt": "For [ACTOR] in [SCENARIO], ..."
}

The volume

Three axes multiply into one coordinate space.

5 change levels

How far you are willing to move: C0 Baseline, C1 Minor Change, C2 Major Change, C3 Structural Change, C4 Paradigm Change.

GET /v1/axes/change-levels

10 internal variables

The clauses of the system sentence: Object, Conditions, Actions, Tools, Resources, Outcomes, Feedback, Constraints, Value, Failure mode.

GET /v1/axes/internal-variables

14 meta-variables

How a clause can vary: Magnitude, Rate, Direction, Scope, Depth, Duration, Frequency, Acceleration, Variability, Detectability, Reversibility, Propagation, Amplification, Accumulation.

GET /v1/axes/meta-variables

5 × 10 × 14 = 700

A coordinate such as LAKA-C3-I07-M08 reads as Structural Change × Feedback × Acceleration. It is a question slot — an analysis prompt — not a guaranteed distinct, feasible, novel or valuable innovation.

Endpoints

22 routes across six groups.

Full parameters and response shapes are in the API reference.

GET/v1Service index: version, counts and the endpoint listing.

GET/v1/healthLiveness check. Returns the loaded coordinate count.

GET/llms.txtPlain-text agent guide: what the grammar is, every route, and the typical flow.

GET/v1/axesAll three axes in one document, with the origin note and the illustrative-state warning.

GET/v1/axes/{axis}One axis on its own.

GET/v1/gridThe 50-cell grid pairing each change level with each internal variable, and the intent of that pairing.

GET/v1/coordinates/{code}Resolve a full coordinate into its three axis definitions, the grid intent and the prompt stem.

GET/v1/promptsSearch and page the 700 prompt stems. Filters combine with AND.

GET/v1/prompts/{id}One prompt stem by coordinate id.

GET/v1/modesThe six operating modes plus the overview table as Markdown.

GET/v1/modes/{id}One mode playbook as Markdown.

GET/v1/operatorsThe seventeen operators with definition, input, output, guardrail and example.

GET/v1/operators/{name}One operator. Case-insensitive.

GET/v1/coreThe seven core framework extracts, listed by id and title.

GET/v1/core/{id}One core extract as Markdown.

GET/v1/templatesThe five working templates, listed by id and title.

GET/v1/templates/{id}One template as Markdown.

GET/v1/schema/runThe JSON Schema (2020-12) a run document must satisfy.

GET/v1/schema/blank-runAn empty, schema-valid run document to fill in.

GET/v1/scoringThe seven weighted criteria, the four hard gates, the rating anchors and the formula.

POST/v1/scoreWeight a set of 0–4 ratings into a 0–100 comparison index, with the per-criterion contribution shown.

POST/v1/validateValidate a run document against the run schema. Returns every error with its JSON path.

Honesty

What this is not.

Coordinates are questions

An unfilled cell does not establish novelty or demand. Filling one produces an observation, assumption, hypothesis or documented not-applicable result — nothing stronger.

State ladders are vocabularies

The illustrative states are not calibrated ordinal scales and do not map onto change levels.

The score is an index

/v1/score returns an analyst-defined comparison number from 0 to 100. It is not a success probability.

DECODE is not proof

A decoded explanation of a system is not proof of intent or wrongdoing.