dcp-wrap

Convert JSON to DCP positional-array format — fewer tokens, same accuracy.

DCP strips repeated keys from structured data. Instead of sending {"endpoint":"/v1/users","method":"GET","status":200} per record, DCP declares the schema once and writes values by position:

["$S","api-response:v1","endpoint","method","status","latency_ms"]
["/v1/users","GET",200,42]
["/v1/orders","POST",201,187]

40–60% token reduction when feeding data to LLMs. Zero accuracy cost. See benchmark.

npm install dcp-wrap

Or use directly:

npx dcp-wrap

cat api-response.json | npx dcp-wrap init api-response

Output:

Schema: api-response:v1
Fields: 4

  endpoint: string (source: endpoint, unique: 4/4)
  method: string (source: method, unique: 2/4)  [enum(2)]
  status: number (source: status, unique: 2/4)
  latency_ms: number (source: latency_ms, unique: 4/4)

Saved: dcp-schemas/api-response.v1.json
Saved: dcp-schemas/api-response.v1.mapping.json

The generator infers field types, detects enums, numeric ranges, and orders fields by DCP convention (identifiers → classifiers → numerics → text).

cat data.json | npx dcp-wrap encode --schema dcp-schemas/api-response.v1.json

Output:

["$S","api-response:v1","endpoint","method","status","latency_ms"]
["/v1/users","GET",200,42]
["/v1/orders","POST",201,187]
["/v1/auth","POST",200,95]

npx dcp-wrap inspect dcp-schemas/api-response.v1.json

For known structures where you define the schema inline:

import { dcpEncode } from "dcp-wrap";

const dcp = dcpEncode(results, {
  id: "engram-recall:v1",
  fields: ["id", "relevance", "summary", "tags", "hitCount", "weight