Dify
Dify is an open-source LLM application development platform that provides orchestration, RAG pipelines, model governance, and custom AI agent workflows. By connecting PostQode to Dify, you can leverage centralized model routing, enterprise security controls, and custom knowledge pipelines.
Website: https://dify.ai/
Getting an API Key
- Log in to your Dify Console or self-hosted Dify instance.
- Navigate to your target application (e.g. Agent App or Chatflow).
- Go to Develop ➔ API Access in the left sidebar.
- Generate and copy your API Secret Key (
app-...).
Supported Modes & Capabilities
PostQode connects to Dify through Dify's OpenAI-compatible and Chat API endpoints:
- Agent App Mode: Routes PostQode prompts through Dify-orchestrated agents with custom system instructions, RAG datasets, and tool integrations.
- Model Gateway Mode: Use Dify as an enterprise proxy to govern model access, manage quotas, and capture audit logs.
Configuration in PostQode
- Open PostQode Settings ().
- Select Dify in the API Provider dropdown.
- Enter your Dify API Endpoint (e.g.
https://api.dify.ai/v1for Dify Cloud, orhttp://your-dify-host/v1for self-hosted). - Enter your Dify API Key (
app-...). - (Optional) Specify target Model ID or custom application parameters.
- Click Save Settings to connect.
Tips and Notes
- Self-Hosted Instances: Ensure your self-hosted Dify instance has network accessibility from your development machine. If using Docker, use
http://localhost:5001/v1or your host IP. - Streaming Support: Dify supports Server-Sent Events (SSE) streaming, ensuring smooth real-time unified diff generation in PostQode.