Tool: export_reasoning_audit

Purpose

Export the canonical reasoning payload plus a deterministic machine- or human-readable artifact for downstream agents and compliance review.

Input

  • deliberation_id (required)
  • format (optional): json (default) or markdown
  • include_raw_thoughts (optional bool, default false)

Output

  • deliberation_id
  • audit_id
  • payload: the unchanged canonical JSON audit object
  • format: the selected format
  • content_type: application/json or text/markdown; charset=utf-8
  • artifact: deterministic rendering of the canonical audit snapshot

Includes

  • session, mode, goal, and route metadata
  • verification statuses and evidence sources
  • assumptions, checkpoints, pipeline gates, and blockers
  • explicit heuristic/provider provenance
  • graph metadata; raw node content only when include_raw_thoughts=true

Markdown tables escape pipes, line breaks, HTML-significant characters, and active Markdown delimiters for links, images, emphasis, code, and strikethrough. Map-derived sections are sorted for stable output. JSON remains the compatibility default, and the canonical payload, audit_id, and deliberation_id fields remain available in both formats.