README

An OpenAI-compatible REST API gateway and reverse adapter for Agentoom.

Important

Post-Splitting Versioning & Pre-Release Notice (v0.8.0) Version 0.8.0 represents the initial standalone release following the modular package extraction and architectural splitting from the Agentoom core monorepo. This package is currently in a pre-release state undergoing integration testing and requires appropriate validation and test coverage across target environments before being designated as production-stable (1.0.0).

🌟 Overview

The agentoom/core-openapi-v1 package exposes standard OpenAI v1 endpoints (/v1/chat/completions, /v1/models) backed by Agentoom's agent network. Any application, SDK, or developer tool designed to talk to OpenAI (such as the official OpenAI Python/Node SDKs, LangChain, LlamaIndex, Cursor, Continue, or LibreChat) can use Agentoom as a direct drop-in replacement by simply altering the base_url and api_key.

Package Name: agentoom/core-openapi-v1
Version: 0.8.0 (Pre-release / Post-split testing)
Namespace: Agentoom\OpenApiV1\
Service Provider: Agentoom\OpenApiV1\OpenApiV1ServiceProvider

🚀 Key Features

1. OpenAI v1 Compatible Endpoints

Chat Completions (POST /v1/chat/completions): Accepts standard OpenAI chat completion JSON payloads (messages, model, temperature, stream, etc.) and executes them via the resolved Agentoom agent.
Model Discovery (GET /v1/models): Lists all available active Agentoom agents mapped as OpenAI model identifiers (e.g. model: "support-agent").

2. High-Performance SSE Streaming

Server-Sent Events (SSE): Full support for stream: true streaming responses.
Transformers: Emits OpenAI-compliant data: {"id": "...", "choices": [{"delta": {"content": "..."}}]} chunks in real time as tokens arrive (StreamTransformer, StreamingRestorer).

3. Stateful Conversation Anchoring

Stateless-to-Stateful Bridge: OpenAI requests are stateless, but Agentoom agents benefit from stateful memory.
Anchor Resolver: ConversationAnchorResolver uses client-provided headers or user identifiers to anchor requests to persistent Agentoom AgentConversation threads, allowing agents to retain long-term context across separate API calls.

4. Enterprise Security & Token Management

API Keys (OpenApiToken): Bearer token authentication with granular permissions.
Agent Scoping: Restrict specific API tokens to only invoke designated agents.
Rate Limiting & Revocation: Enforce request rate limits and immediately revoke compromised keys.

5. Integrated Privacy Protection

Privacy Guard (ConversationPrivacyGuard): Intercepts inbound prompts and outbound responses through Agentoom's privacy pipeline (agentoom/core-privacy), stripping sensitive PII before completion.

6. Administrative Management

Livewire Admin Interface: Complete dashboard for managing API keys, inspecting call volumes, and viewing client connection logs.
Multi-Language Support: Fully localized in English (en), German (de), Spanish (es), French (fr), and Italian (it).

📁 Package Structure

core-openapi-v1/
├── config/
│   └── openapi-v1.php              # Route prefix, default settings, rate limits
├── database/
│   ├── factories/                  # OpenApiTokenFactory
│   └── migrations/                 # Table schema for openapi_tokens
├── resources/
│   ├── lang/                       # Translations (de, en, es, fr, it)
│   └── views/livewire/admin/       # Livewire token management views
├── routes/
│   ├── admin.php                   # Token admin routes
│   └── api.php                     # /v1/chat/completions and /v1/models routes
└── src/
    ├── DTOs/                       # ChatCompletionRequest, AgentExecutionInput, Message
    ├── Http/
    │   ├── Controllers/            # ChatCompletionController, ModelController
    │   └── Middleware/             # AuthenticateOpenApiToken
    ├── Models/
    │   └── OpenApiToken.php        # Token model
    ├── Services/                   # DynamicAgentExecutionAdapter, AnchorResolver, PrivacyGuard
    ├── Transformers/               # ChatCompletionTransformer, StreamTransformer
    └── OpenApiV1ServiceProvider.php

💻 Usage Example

Using the official OpenAI Python SDK:

from openai import OpenAI

client = OpenAI(
    base_url="https://your-agentoom-domain.com/v1",
    api_key="agt_live_abcdef1234567890..."
)

response = client.chat.completions.create(
    model="customer-support-agent",
    messages=[
        {"role": "user", "content": "Hello, how can I track my shipment?"}
    ],
    stream=True
)

for chunk in response:
    content = chunk.choices[0].delta.content or ""
    print(content, end="", flush=True)

agentoom / core-openapi-v1

Package info

Security

Statistics