Stream chat responses for cluster troubleshooting

Runs one AI chat turn about the cluster and streams the reply as server-sent events. Answers 404 when the cluster is not a live cluster of the organisation. Deprecated: responses carry Deprecation and Sunset headers, and durable chat sessions (/api/v1/chat/sessions) are the successor.

POST /api/v1/org/clusters/{cluster_id}/kubernetes/chat

Stream chat responses for cluster troubleshooting

Python

import requests

url = "https://platform.ankra.app/api/v1/org/clusters/{cluster_id}/kubernetes/chat"

payload = {
    "query": "<string>",
    "addon_name": "<string>"
}
headers = {
    "Authorization": "Bearer <token>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

422

{
  "detail": [
    {
      "loc": [
        "<string>"
      ],
      "msg": "<string>",
      "type": "<string>"
    }
  ]
}

Authorizations

Authorization · string · header · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Headers

authorization · string | null

x-ankra-organisation-id · string | null

Path Parameters

cluster_id · string<uuid4> · required

Body

application/json

query · string · required

addon_name · string | null

chart_name · string | null

chart_version · string | null

compliance_focus · ComplianceFocus · object | null

Show child attributes

conversation_history · ConversationMessage · object[] | null

Show child attributes

conversation_id · string | null

current_values · Current Values · object | null

default_values · Default Values · object | null

draft_id · string | null

history_forked · boolean | null

When true, this turn forks the conversation at the end of the supplied conversation_history (edit-and-resend): the server fills in the turns before it from the stored transcript and never the branch after it.

idempotency_key · string | null

images · ImageAttachment · object[] | null

Show child attributes

include_logs · boolean · default: true

interaction_mode · enum<string> · default: agentic

Available options: ask, agentic, plan

is_config_page · boolean · default: false

launch_plan_approval_id · string<uuid> | null

The plan-level Launch Plan approval this turn acts under. A confirmation-gated write the approval covers is claimed under it and runs instead of asking for confirmation; anything it does not cover, and any call after the approval expired, was revoked or the plan changed, asks for confirmation as usual. Only the person who approved can act under it.

max_log_lines · integer · default: 50

model_mode · string | null

namespace · string | null

page_context · PageContext · object | null

Show child attributes

rag_context · RAGContext · object | null

Show child attributes

regenerate_seed · string | null

repository_name · string | null

resource_name · string | null

resource_type · string | null

supersedes_plan_id · string | null

surface · enum<string> | null

The client rendering this turn. "web" is the native web chat, which can show a secure secret-entry form, so the session runner offers request_secret_input on the turn. Absent or empty keeps the default behaviour (no form is ever requested).

Available options: web

Response

200

application/json

Streaming chat response