@nexalab/agent-sdk 0.1.0 → 0.1.5

package/README.md

@@ -505,14 +505,36 @@ Recommended production wiring:
 - Use `pnpm harness` as the production readiness gate.
 - Use `pnpm smoke` and `pnpm pack:check` before publishing.
 
-## Is It Frontier?
+```md
+## Frontier-ready architecture
 
-Yes, this is now a frontier-style agent SDK architecture: native tool calling, structured streaming, durable storage adapters, policy/approvals, evals, telemetry, replay, workflows, routing, multi-agent coordination, vector memory, queue contracts, package smoke checks, and a production harness.
+Nexa Agent SDK is designed with a frontier-style agent architecture: native tool calling, structured streaming, durable storage adapters, policy and approvals, evals, telemetry, replay, workflows, routing, multi-agent coordination, vector memory, queue contracts, package smoke checks, and an extensible `ModelHarness` contract.
 
-The SDK intentionally stays dependency-light. Production apps still provide their actual infrastructure drivers and deployment policy through adapters.
+The SDK remains dependency-light. Production applications provide their own infrastructure drivers and deployment policies through adapters.
 
-## Open-core architecture
+## Extensible by design
 
-This package is designed as the fully-capable public half of an open-core split — see [docs/architecture/open-core-audit.md](./docs/architecture/open-core-audit.md), [open-core-target.md](./docs/architecture/open-core-target.md), and [open-core-migration.md](./docs/architecture/open-core-migration.md) for the module-by-module classification, target architecture, and phased plan. In short: everything in this repository is public and production-capable; strategic/proprietary intelligence (e.g. a future `FrontierHarness`) is designed to live in a separate private repository that *implements* this SDK's public contracts (`ModelHarness`, `ModelRoute`, `ContextEngineConfig`, ...) rather than forking or being vendored into it.
+Nexa is built around stable contracts instead of hardcoded orchestration strategies.
 
-`pnpm verify:boundaries`, `pnpm verify:secrets`, and `pnpm verify:publish-safety` (all folded into `pnpm prepublish:check`) guard against private code or secrets ever entering an npm-published tarball from this repo.
+Developers can extend or replace major runtime behaviors through interfaces for:
+
+- model harnesses
+- providers
+- tools
+- context
+- memory
+- storage
+- routing
+- telemetry
+- evals
+
+This makes it possible to build custom agent architectures without forking the core SDK.
+
+## Release safety
+
+The repository includes automated checks for package boundaries, accidental credential exposure, and npm tarball contents:
+
+```bash
+pnpm verify:boundaries
+pnpm verify:secrets
+pnpm verify:publish-safety

package/package.json

@@ -1,6 +1,10 @@
 {
   "name": "@nexalab/agent-sdk",
-  "version": "0.1.0",
+  "version": "0.1.5",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/pixelitylab/nexalab-agentsdk.git"
+  },
   "private": false,
   "description": "Reusable agentic SDK extracted from Nexa Agent OS.",
   "license": "Apache-2.0",
@@ -8,7 +12,7 @@
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
   "bin": {
-    "nexa": "./bin/nexa.mjs"
+    "nexa": "bin/nexa.mjs"
   },
   "exports": {
     ".": {