sre-task-refinement

Task refinement applies SRE-level review to examine granularity, implementability, success criteria, dependencies, safety standards, and edge cases. It identifies hidden failure modes and vague instructions that could cause production problems, ensuring junior engineers can execute tasks without ambiguity.

Actionable plans require clear granularity (4-8 hour subtasks), 3+ measurable success criteria, explicit dependency structure, complete safety standards, edge-case coverage, and meaningful tests. Task refinement enforces these eight categories systematically to eliminate implementation blockers.

Tasks exceeding 16 hours should be broken into 4-8 hour subtasks with self-contained designs. This granularity prevents implementation paralysis, reduces risk of hidden dependencies, and enables junior engineers to complete work without extended context switching.

SRE-level review systematically examines dependencies, safety assumptions, and boundary conditions to surface hidden failure modes. Red flag detection catches antipatterns and incomplete specifications that testing alone might miss, preventing production incidents.

Measurable success criteria are quantifiable, observable outcomes tied to task completion. Task refinement requires 3+ per task to eliminate subjective interpretation and ensure tests validate actual functionality rather than ambiguous intent.