# Connect OpenAI Codex to Ishara

Connect OpenAI Codex (the CLI, the IDE extension and the ChatGPT desktop app) to Ishara’s MCP server in config.toml.

- Server: `https://useishara.com/mcp`
- Where it goes: `~/.codex/config.toml`, shared by the CLI, the IDE extension and the ChatGPT desktop app
- Transport: Streamable HTTP, built in
- Authentication: a bearer token in the `Authorization` header

## Steps

1. In Ishara, open **Agents**, create a token under **New token** and copy it (it’s shown once).
2. Add Ishara to OpenAI Codex, with your token in place of `ishara_…`:

   ~/.codex/config.toml:

```toml
[mcp_servers.ishara]
url = "https://useishara.com/mcp"
bearer_token_env_var = "ISHARA_TOKEN"
```

   Then set `ISHARA_TOKEN` to your token in your shell profile. Codex adds the “Bearer” part itself.

   Or with the token in the file:

```
[mcp_servers.ishara]
url = "https://useishara.com/mcp"
http_headers = { Authorization = "Bearer ishara_…" }
```

3. Type `/mcp` in Codex, or run `codex mcp list`: Ishara is listed with its tools.
4. Ask your agent: “Apply the open Ishara notes.” It lists them, applies each edit in your code, and marks each note done with a reply the reviewer sees on the site.

## Good to know

- `codex mcp add` has no header option, so add the token line by hand.
- A project’s own `.codex/config.toml` is read only in projects you’ve marked as trusted.

Checked against the official docs on 2026-11-12: [Codex: MCP](https://learn.chatgpt.com/docs/extend/mcp.md), [Codex: configuration reference](https://learn.chatgpt.com/docs/config-file/config-reference.md).
