skill-evolution

To land skill improvement proposals with human approval, you apply this skill to pending proposals, draft the exact diff, and present it for an explicit human vote before applying changes to EN and ZH skill trees.

Triageing captured skill proposals for governance involves automating the routing of improvement signals to the right human gate, drafting the exact diff, and maintaining an explicit audit trail with verifiable landing records.

Cross-tree synchronization works by applying approved changes to both EN and ZH skill trees, enforcing governance by validating modifications through the shared skill validation flow and recording outcomes under each proposal.

Preventing silent self-modifications of skills requires a triage workflow that gates the evolution process, ensuring changes are vetted and landed only after an explicit human approval and validation.

You should use a human-approval workflow for skill evolution when you need to ensure captured improvements are vetted, prevent unauthorized self-modifications, and enforce accountability through recorded outcomes and status updates.

Yes, you need a shared validation flow to land changes across skill trees, as it validates the drafted diffs and enforces governance by updating the proposal status to landed or rejected after synchronization.