@tekmemo/openai
v0.2.0
Published
Production-ready OpenAI embedding adapter for TekMemo.
Readme
@tekmemo/openai
Purpose
OpenAI embeddings. OpenAI embedding adapter for TekMemo recall pipelines.
Install
pnpm add @tekmemo/openaiQuick start
import { createOpenAIEmbedder } from "@tekmemo/openai";
const embedder = createOpenAIEmbedder({ apiKey: process.env.OPENAI_API_KEY! });Boundary
This package owns its package-level contract only. It does not own TekMemo Cloud billing, dashboards, tenancy, hosted database storage, or provider secrets unless explicitly stated by its package name.
For hosted memory, use @tekmemo/cloud-client. For local file-backed memory, use tekmemo with @tekmemo/fs. For MCP tools, use @tekmemo/mcp-server.
Scripts
pnpm --filter @tekmemo/openai typecheck
pnpm --filter @tekmemo/openai test:run
pnpm --filter @tekmemo/openai build
pnpm --filter @tekmemo/openai lint:packageDocs
- Package docs: https://docs.tekmemo.dev/packages/
- Examples: https://docs.tekmemo.dev/examples/
- Repository: https://github.com/tekbreed/tekmemo
Publishing metadata
- npm package:
@tekmemo/openai - publish visibility: public
- runtime format: dual ESM/CJS
- ESM output:
dist/**/*.mjs+dist/**/*.d.mts - CJS output:
dist/**/*.cjs+dist/**/*.d.cts - package contents:
distandREADME.md - package boundary: hosted cloud calls must go through
@tekmemo/cloud-clientunless this package is@tekmemo/cloud-clientitself.
Publish readiness
Before publishing this package, run:
pnpm --filter @tekmemo/openai release:checkThe package-level check builds dist/, runs TypeScript and tests, runs publint, and performs npm pack --dry-run. Publish from CI with Changesets and npm trusted publishing/provenance after the root release preflight passes.
License
MIT.
