dsh-plugin-devin-bridge

Configuration

Edit cordis.patch.yml in the repository (or the copy installed in your profile) and fill in your Devin session token:

- insert:
    - id: devin-bridge
      name: dsh-plugin-devin-bridge
      config:
        baseUrl: "https://server.codeium.com"
        token: "devin-session-token$..."   # required
        model: "glm-5-2"                    # default model uid
        proxy: ""                           # optional outbound proxy
        forceHttp1: true                    # force HTTP/1.1 to Devin
        apiKey: ""                          # optional /v1/* access key

field	required	description
baseUrl	yes	Devin Connect endpoint, usually https://server.codeium.com.
token	yes	Devin session token, format devin-session-token$.
model	no	Default model uid when the request does not specify one. Defaults to glm-5-2.
proxy	no	Outbound proxy URL. Supports http://, https://, socks5://, socks5h://.
forceHttp1	no	Force HTTP/1.1 to Devin. Defaults to true.
apiKey	no	Protect /v1/* with Authorization: Bearer . Empty = no auth.

Usage

Once the profile is running, any OpenAI-compatible client works:

from openai import OpenAI
client = OpenAI(
    base_url="http://127.0.0.1:/v1",
    api_key="",
)
resp = client.chat.completions.create(
    model="glm-5-2",
    messages=[{"role": "user", "content": "Hello"}],
    stream=True,
)
for chunk in resp:
    print(chunk.choices[0].delta.content or "", end="")

List available models:

curl http://127.0.0.1:/v1/models \
  -H "Authorization: Bearer "