continuous-learning-v2

You can convert coding sessions into reusable skills by enabling observation hooks that extract confidence-scored atomic instincts from your workflow, which the system then evolves into actionable commands or agents.

Project-scoped instincts are isolated, confidence-weighted patterns stored within a specific project to prevent cross-project contamination, while allowing proven universal patterns to be promoted to global instincts for team reuse.

Yes, you need a local environment with Python and shell scripts to run the observer, detect projects, manage atomic instincts, and evolve detected patterns into actionable skills.

Confidence scoring weighs atomic instincts extracted from observed sessions, allowing the system to validate patterns before promoting them from project-scoped isolation into global skills, commands, or agents.

Yes, the system supports import, export, and promotion features that enable sharing confidence-scored instinct patterns across different projects and teams to facilitate cross-project reuse.

The best way to automate pattern extraction is enabling the observation hooks to passively monitor sessions, automatically detecting and scoring atomic instincts without requiring manual intervention.