open-falsification-probe

name	open-falsification-probe
description	Open compact falsification probes from the next-action queue so contradiction-heavy or low-confidence board targets become explicit probe objects.

Eco Open Falsification Probe

Core Goal

• Convert probe-worthy next actions into explicit falsification probes.
• Preserve the target ids, linked evidence refs, and governance follow-up actions.
• Do not select professional analysis, query, transport, representation, or diffusion tools for the council.
• Emit a durable investigation artifact for challenger and moderator workflows.

Triggering Conditions

• A ranked action queue contains contradiction-heavy or low-confidence targets.
• Need explicit probe objects instead of leaving falsification work implicit in notes or challenges.
• Need a bridge artifact between D1 action planning and D2 readiness gating.

Read/Write Contract

• Reads run_dir/investigation/next_actions_<round_id>.json by default.
• If the next-actions artifact is absent, rebuilds probe candidates from the run-local deliberation plane plus analysis-plane-backed coverage context.
• Writes run_dir/investigation/falsification_probes_<round_id>.json by default.

Required Input

• run_dir
• run_id
• round_id
• Optional:
  • next_actions_path
  • board_summary_path
  • board_brief_path
  • coverage_path
  • output_path
  • action_id
  • max_probes
  • max_actions

Output Contract

• status
• summary
• receipt_id
• batch_id
• artifact_refs
• canonical_ids
• warnings
• deliberation_sync
• analysis_sync
• board_handoff
• The emitted artifact also carries normalized D1 trace metadata in action_source, board_state_source, coverage_source, db_path, and observed_inputs, including explicit next_actions_artifact_present handling on both artifact and fallback paths.

References

• ../../docs/openclaw-project-overview.md

Scripts

• scripts/open_falsification_probe.py