Codex | Synthflow

Synthflow AI Codex connector

The Codex connector lets you work with your Synthflow workspace directly from Codex. It connects Codex to the Synthflow MCP server, so you can create and manage agents, place and review calls, build knowledge bases, and run actions without leaving the tool you already use.

This is a good fit when your workflow lives in Codex and you want to manage Synthflow with natural-language prompts. If you prefer in-dashboard assistance with Memory, Learnings, and UI cards, use Aurora instead. For the full list of available tools, see the capabilities table.

Setup instructions

You need a Synthflow API key. The Codex connector uses the Synthflow MCP server at:

URL: https://mcp.synthflow.ai/mcp
Header: Authorization: Bearer <SYNTHFLOW_API_KEY>

Type this prompt into Codex and let it configure the server for you:

Type into Codex

Add the Synthflow MCP server to my Codex configuration.

Add this to `~/.codex/config.toml`:

```toml
[mcp_servers.synthflow]
url = "https://mcp.synthflow.ai/mcp"
bearer_token_env_var = "SYNTHFLOW_API_KEY"
```

Then set the `SYNTHFLOW_API_KEY` environment variable to my Synthflow API key:

```bash
export SYNTHFLOW_API_KEY="<SYNTHFLOW_API_KEY>"
```

Example prompts

Example prompt
List my agents and summarize what each one does.
Create a new inbound sales qualification agent and publish it.
Show me the last 20 calls and flag any with negative sentiment.
Create a knowledge base from these sources, then attach it to my support agent.
Run a simulation where a prospect asks about pricing and implementation time.

Security considerations

Connecting an LLM to operational tools introduces new risks. The primary risk class unique to LLM workflows is prompt injection: untrusted content (like call transcripts, user messages, or knowledge base documents) can include instructions that try to trick the model into taking unintended actions. Most clients support confirm before running tools. Keep that enabled and review tool calls carefully.

FAQ

How is the Codex connector different from Aurora?

Aurora is the in-dashboard agent: one chat thread in the Synthflow UI with Memory, Learnings, and rich UI cards. The Codex connector runs in Codex and authenticates with an API key. Both can manage agents, calls, and knowledge bases.

How is this different from MCP Connectors?

MCP Connectors let a voice agent call tools on an external MCP server during a live call. The Codex connector is the opposite direction: an AI agent in Codex calls tools to manage your Synthflow workspace.

Which API key permissions are required?

Use the same API key you would use for REST. Generate and manage keys in Admin → Workspace Settings → API Keys. See Authentication for setup and key-management practices.

How is model usage billed?

The Codex connector operates within your own Codex account, so model and subscription charges are billed directly by Codex according to your plan. Synthflow bills only for the workspace usage your actions generate, such as calls, under your existing Synthflow plan.