update-parker-skill

Automating system updates safely requires a propose-then-execute workflow that identifies dependencies and propagates changes across related components to prevent drift and broken references.

Dependency management in a context-engineering system is best handled by automatically identifying dependent files and propagating updates cross-references to maintain consistency before changes are executed.

Proposing updates to system docs and rubrics without drift involves submitting a description of the change for human approval, ensuring global changes are reviewed before execution.

Global system changes require human approval before execution to ensure proper change management and minimize the risk of errors like missing impacts or broken references.

Updating context-engineering prompts causes missing impacts when cross-reference propagation fails to identify and update dependent files, leading to system inconsistency and drift.

The update-parker-skill workflow supports custom change descriptions passed as arguments when proposing updates to ensure the system understands the intended modifications before approval.