Self-Evolution Architect
Ki · Reasoning
The Problem
The project is complete. The team moves on. No retrospective, no process audit, no record of what worked or what burned time. The same mistakes will repeat on the next project because nobody stopped to extract the lessons.
The Operation
This cognitive operation forces the model to identify every reasoning step taken since the last checkpoint and enumerate them in sequence. Audit each step for efficiency, asking whether it advanced the goal or introduced noise. Classify steps as productive, neutral, or wasteful based on their contribution to the conclusion. Extract lessons by comparing actual reasoning path against an optimal hypothetical path. Verify that at least one actionable improvement is identified. The reasoning applies a formal computation: divergence = count(actual steps != planned steps) / total steps. If it detects skip retrospection after completed reasoning chain, it halts and corrects.
The Structure
This ability runs on a chain of custody audit that traces the provenance of every claim. The procedure repeats until diminishing returns trigger an exit.
If the task concludes without a reflection on what worked, what failed, and what should change next time, post-task debrief was omitted.
Haki · Reasoning-Multi
Cross-Domain Suppression
In Haki mode, the API retrieves the primary ability first, then fans out to three synergy roles that compound its reasoning.
When retrieved in Haki mode, the primary ability is augmented with failure guards extracted from 3 abilities in different cognitive domains. Each guard blocks a specific reasoning failure the primary alone wouldn't catch. A self-check forces verification before output. The result is cross-domain coverage that no single ability can reach alone.