synthesize

Runbook generation from a planning brief is automated by synthesizing formal, validated step sequences using a canonical vocabulary. The process enforces frontmatter constraints, mandatory tail requirements, and per-step artifacts to produce a production-grade, auditable plan.

A canonical vocabulary runbook for engineering orchestration is a structured plan that uses only predefined step IDs from a validated reference file. It ensures traceable, compliant step sequences required for complex tasks like feature work or migrations.

Validating step IDs in a synthesized runbook involves checking each step definition against a dedicated reference vocabulary file. This ensures all steps, frontmatter fields, and tail requirements conform to canonical constraints before execution.

You can include custom scripts, references, and assets in an automated runbook by directing the skill to attach optional resources. This supports contract pairings and complex release artifacts alongside the validated step sequences.

Automated runbook synthesis supports human-in-the-loop validation by producing per-step HITL signals within the generated plan. This allows teams to iterate on the production-grade skeleton while maintaining formal validation and traceability.

The limitation of using a canonical vocabulary for runbook synthesis is that all step IDs must strictly match the predefined reference file. Runbooks requiring steps outside this validated vocabulary will fail validation constraints and cannot be generated.