orchestrating-workflows

Orchestrating SDLC workflows for plugins involves sequencing sub-skill invocations to drive feature, chore, and bug tasks while managing persistent state across pause and resume points. You start by calling the orchestrator with a feature title or issue reference.

State management across pause points works by maintaining persistent state data that survives sub-agent forking and phase transitions. This allows the orchestrator to resume workflow execution seamlessly and track progress through per-phase loops.

Yes, issue-tracking integration is supported natively. You can begin orchestration by passing an issue reference to the orchestrator, which then coordinates the workflow phases and logs progress through audit trails.

Adaptive model selection dynamically resolves processing tiers across workflow forks. The orchestrator manages this feedback loop via audit-trail logging, ensuring appropriate model allocation for different phases of the SDLC workflow.

Yes, jq is a required dependency. You need this command-line JSON processor installed in your environment to support the orchestrator's state management and data parsing operations.

This orchestrator is designed specifically for Claude Code plugins and relies on sub-skill coordination. It may not suit environments lacking compatible sub-skills or where native plugin architecture is unsupported.