Claude Code (Credentials) icon

Claude Code (Credentials)

Use Claude Code SDK to execute AI-powered coding tasks with authentication from n8n credentials

Actions2

Continue

AI-generated

Summary

Continue a previous Claude Code conversation using a session ID, sending a new prompt that extends the dialogue, and receive the assistant's structured or textual response.

Inputs

  • Session ID (required) — The session ID from a prior Query operation to continue that conversation.
  • Prompt (required) — The prompt or message to send to Claude Code as the next turn in the conversation.
  • Model — Select the Claude model to use ('Sonnet' for fast and efficient or 'Opus' for most capable). Defaults to 'Sonnet'.
  • Max Turns — Maximum number of conversation turns allowed before stopping. Defaults to 25.
  • Timeout — Maximum time in seconds to wait for response before aborting. Defaults to 300 seconds.
  • Project Path — Working directory path for Claude Code operations; if omitted uses stored path from session.
  • Output Format — Format of returned data: structured detailed output, raw messages array, or final text only. Defaults to 'structured'.
  • Allowed Tools — List of Claude Code tools allowed during the conversation; default includes WebFetch, TodoWrite, WebSearch, Task.
  • Disallowed Tools — List of Claude Code tools to explicitly block even if allowed.
  • Additional Options — Optional settings including debug logging, permission handling mode, and system prompt additions.

Output shape

a single JSON object containing response details

The output JSON includes the session ID, success status, duration, and conversation messages or result text depending on output format. Secrets and sensitive info are redacted. Errors are reported in JSON with error type. Supports multiple input items with paired outputs.

Links

Discussion