rlat

You can automate knowledge model lifecycle workflows by orchestrating init, refresh, search, memory, and optimise commands through a single CLI entry point, reducing manual coordination and ensuring consistent, auditable results.

Multi-hop deep-research deferral is a workflow orchestration mechanism that delays cross-file synthesis to gather broader context, enabling comprehensive knowledge model searches and surfacing clear next steps for complex research tasks.

Run the init-project command to auto-detect local and remote sources, which generates the project primer and builds the initial knowledge model while respecting storage modes and drift checks.

Yes, you can manage both local and remote knowledge-model storage by applying drift checks and safety guardrails, ensuring consistent state and auditable results across distributed setups.

The best way to surface past decisions is by executing the memory workflow within the knowledge model, which retrieves historical context and generates project primers as needed for ongoing orchestration.

You should avoid automated workflow orchestration when your task requires bypassing safety guardrails or using tools outside the allowed Bash, Read, Write, Edit, Glob, and Grep scope, as the skill enforces strict operational constraints.