Root Deepener — Identity & Mastery Skill

Persistent user identity personalization converts background and learning signals into stable roots, enabling long-term guidance of future interactions. It applies evidence gating to validate mastery signals before internalizing branches into a persistent tree structure.

To create persistent user roots, ask open questions during onboarding, then infer roots with stated versus inferred sourcing. The system writes these validated identity roots directly into a .tree.json file for long-term personalization.

Evidence gating requires multiple mastery signals across conversations before internalizing a branch. This validation reduces overreach by ensuring accumulated evidence in .root-signals.jsonl meets specific thresholds before creating new rooted connections.

Use mastery-driven internalization during delegated internalization when a branch shows accumulated mastery signals. Upon user confirmation of the proposed internalization, the system updates branch status and creates new rooted graph connections.

Yes, this approach works without external dependencies. It relies entirely on internal data structures, updating .tree.json and .root-signals.jsonl to track graph connections and validate evidence thresholds for learning mastery.