@open-multi-agent/core 1.4.0 → 1.4.2

package/docs/providers/minimax.md

@@ -0,0 +1,75 @@
+# MiniMax setup guide
+
+<img src="../../.github/brand/minimax-banner.png" alt="MiniMax" width="600">
+
+MiniMax M2.7 can be used in OMA's TypeScript multi-agent workflows with a simple provider config.
+
+## Community offer
+
+OMA users get **12% off** the MiniMax Token Plan. Valid until **2026-06-30**.
+
+| Region | Link |
+|--------|------|
+| Global | [platform.minimax.io](https://platform.minimax.io/subscribe/coding-plan?code=6ZoOY13DDV&source=link) |
+| China | [platform.minimaxi.com](https://platform.minimaxi.com/subscribe/token-plan?code=98qruMqQhL&source=link) |
+
+This is a limited-time community offer, not a paid endorsement.
+
+## About MiniMax M2.7
+
+MiniMax describes M2.7 as:
+
+> MiniMax M2.7 is MiniMax's first self-evolution model, capable of autonomously building complex Agent Harnesses and completing high-complexity productivity tasks via Agent Teams, advanced Skills, and Tool Search Tool. It excels in software engineering, end-to-end project delivery, and office scenarios, with stable task execution, environment interaction, and identity-preserving capabilities.
+
+MiniMax 对 M2.7 的描述：
+
+> MiniMax M2.7 是 MiniMax 首个深度参与自我迭代的模型，可自主构建复杂 Agent Harness，并基于 Agent Teams、复杂 Skills、Tool Search Tool 等能力完成高复杂度生产力任务；其在软件工程、端到端项目交付及办公场景中表现优异，多项评测接近行业领先水平，同时具备稳定的复杂任务执行、环境交互能力以及良好的情商与身份保持能力。
+
+## Setup
+
+### Environment variables
+
+```bash
+export MINIMAX_API_KEY=your-api-key
+```
+
+The adapter defaults to the global endpoint (`https://api.minimax.io/v1`). China users should override the base URL:
+
+```bash
+export MINIMAX_BASE_URL=https://api.minimaxi.com/v1
+```
+
+### Agent config
+
+```typescript
+const agent: AgentConfig = {
+  name: 'my-agent',
+  provider: 'minimax',
+  model: 'MiniMax-M2.7',
+  systemPrompt: 'You are a helpful assistant.',
+}
+```
+
+Full example:
+
+```typescript
+import { OpenMultiAgent, type AgentConfig } from '@open-multi-agent/core'
+
+const agent: AgentConfig = {
+  name: 'analyst',
+  provider: 'minimax',
+  model: 'MiniMax-M2.7',
+  systemPrompt: 'Analyze data and produce concise reports.',
+  tools: ['bash', 'file_read', 'file_write'],
+}
+
+const orchestrator = new OpenMultiAgent()
+const result = await orchestrator.runAgent(agent, 'Summarize /tmp/report.csv')
+console.log(result.output)
+```
+
+## Disclosure
+
+- This is a limited-time community offer valid through 2026-06-30.
+- Listings are not paid endorsements.
+- Some provider offers may include referral credits that help maintain the project.

package/docs/providers.md

@@ -44,6 +44,8 @@ No bundled shortcut is needed when a server speaks OpenAI Chat Completions. Use
 | OpenRouter | `provider: 'openai'` + `baseURL: 'https://openrouter.ai/api/v1'` + `apiKey` | `OPENROUTER_API_KEY` | `openai/gpt-4o-mini` | |
 | Groq | `provider: 'openai'` + `baseURL: 'https://api.groq.com/openai/v1'` | `GROQ_API_KEY` | `llama-3.3-70b-versatile` | |
 | Mistral | `provider: 'openai'` + `baseURL: 'https://api.mistral.ai/v1'` | `MISTRAL_API_KEY` | `mistral-large-latest` | See [`providers/mistral`](../examples/providers/mistral.ts). |
+| Zhipu GLM | `provider: 'openai'` + `baseURL: 'https://open.bigmodel.cn/api/paas/v4'` | `ZHIPU_API_KEY` | `glm-4-plus` | See [`providers/zhipu`](../examples/providers/zhipu.ts). |
+| Doubao (ByteDance) | `provider: 'openai'` + `baseURL: 'https://ark.cn-beijing.volces.com/api/v3'` | `ARK_API_KEY` | `doubao-1-5-pro-32k-250115` | See [`providers/doubao`](../examples/providers/doubao.ts). |
 
 Other services can be connected the same way if they implement the OpenAI Chat Completions API, but they are not listed as verified providers here. For services where the key is not `OPENAI_API_KEY`, pass it explicitly via `apiKey`; otherwise the `openai` adapter falls back to `OPENAI_API_KEY`.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@open-multi-agent/core",
-  "version": "1.4.0",
+  "version": "1.4.2",
   "description": "TypeScript multi-agent framework — one runTeam() call from goal to result. Auto task decomposition, parallel execution. 3 dependencies, deploys anywhere Node.js runs.",
   "files": [
     "dist",
@@ -25,6 +25,10 @@
     "./mcp": {
       "types": "./dist/mcp.d.ts",
       "import": "./dist/mcp.js"
+    },
+    "./ai-sdk": {
+      "types": "./dist/ai-sdk.d.ts",
+      "import": "./dist/ai-sdk.js"
     }
   },
   "scripts": {
@@ -70,7 +74,8 @@
   "peerDependencies": {
     "@aws-sdk/client-bedrock-runtime": "^3.700.0",
     "@google/genai": "^1.48.0",
-    "@modelcontextprotocol/sdk": "^1.18.0"
+    "@modelcontextprotocol/sdk": "^1.18.0",
+    "ai": "^5.0.0"
   },
   "peerDependenciesMeta": {
     "@aws-sdk/client-bedrock-runtime": {
@@ -81,14 +86,20 @@
     },
     "@modelcontextprotocol/sdk": {
       "optional": true
+    },
+    "ai": {
+      "optional": true
     }
   },
   "devDependencies": {
+    "@ai-sdk/openai": "^2.0.106",
+    "@ai-sdk/provider": "^2.0.3",
     "@aws-sdk/client-bedrock-runtime": "^3.1040.0",
     "@google/genai": "^1.48.0",
     "@modelcontextprotocol/sdk": "^1.18.0",
     "@types/node": "^22.0.0",
     "@vitest/coverage-v8": "^2.1.9",
+    "ai": "^5.0.188",
     "tsx": "^4.21.0",
     "typescript": "^5.6.0",
     "vitest": "^2.1.0"