Handle tool calls from the model. Args: response: LLM response containing tool calls user_input: Ori

Check if AWS credentials are available

Check if AWS credentials are available

Check if AWS credentials are available

Check if AWS credentials are available

Check if Slack webhook URL is available

Check if Slack webhook URL is available

Check if Slack webhook URL is available

Check if Slack webhook URL is available

Placeholder RDS reboot for learning/testing

Placeholder RDS reboot for learning/testing

Placeholder RDS reboot for learning/testing

Placeholder RDS reboot for learning/testing

Real AWS RDS reboot using boto3

Real AWS RDS reboot using boto3

Real AWS RDS reboot using boto3

Real AWS RDS reboot using boto3

Placeholder Slack notification for learning/testing

Placeholder Slack notification for learning/testing

Placeholder Slack notification for learning/testing

Placeholder Slack notification for learning/testing

Send real Slack notification via webhook

Send real Slack notification via webhook

Send real Slack notification via webhook

Send real Slack notification via webhook

One-line preview of a tool result for the progress panel.

One-line preview of a tool result for the progress panel.

One-line preview of a tool result for the progress panel.

One-line preview of a tool result for the progress panel.

Execute tool calls in a loop until the LLM returns a text-only response or we hit the iteration limit.

Execute tool calls in a loop until the LLM returns a text-only response or we hit the iteration limit.

Execute tool calls in a loop until the LLM returns a text-only response or we hit the iteration limit.

Execute tool calls in a loop until the LLM returns a text-only response or we hit the iteration limit.

Save a new incident summary and return it.

Save a new incident summary and return it.

Save a new incident summary and return it.

Return all entries within `window_seconds` of an anchor entry. Used to answer "what else was happening at the same time as this

Render a chronological text timeline with cross-source clusters marked. Format: TIMELINE (last 30m, 47 entries across 3 sources)

Clear the conversation history

Convert Streamlit messages to LangChain message format

Convert Streamlit messages to LangChain message format

Display all chat messages from history

Display all chat messages from history

Render all past messages, including collapsed thinking steps.

Render all past messages, including collapsed thinking steps.

Render all past messages, including collapsed thinking steps.

Render all past messages, including collapsed thinking steps.

Display sidebar with information and controls

Display sidebar with information and controls

executeCreateConfigMap handles ConfigMap creation/update

executeCreateConfigMap handles ConfigMap creation/update

executeDeletePod handles pod deletion

executeDeletePod handles pod deletion

executeGetPodLogs handles log retrieval

executeGetPodLogs handles log retrieval

executeListPods handles listing pods in a namespace

executeListPods handles listing pods in a namespace

executeRestartDeployment handles deployment restarts

executeRestartDeployment handles deployment restarts

executeScaleDeployment handles deployment scaling

executeScaleDeployment handles deployment scaling

formatToolError formats tool execution errors

formatToolError formats tool execution errors

formatToolResult formats successful tool execution results

formatToolResult formats successful tool execution results

* The extension does not talk to the MCP server directly. Instead it routes * user intent through GitHub Copilot Chat, which already has the * `k8s-