plan-review-output-format

To standardize JSON output for plan review agents, enforce a predictable schema with top-level fields like lens, summary, strengths, and findings. This ensures automated orchestrators can parse agent outputs reliably into structured reports.

The required JSON schema for multi-lens plan review pipelines includes top-level fields for lens, summary, strengths, and findings. Each finding object must contain keys for severity, confidence, location, and title to ensure deterministic merging.

Format automated agent findings for CI gating by returning a single JSON block with no surrounding prose. Use Unicode emoji prefixes for severity levels within each finding body to maintain machine-readable structure.

Yes, the plan review format requires specific per-finding keys including severity and confidence. It also mandates location and title fields, alongside Unicode emoji prefixes for finding bodies, to ensure automated tooling behaves deterministically.

You need a standardized JSON output format for orchestrators when merging findings from multiple lens-specific agents in automated review workflows. It removes ambiguous narrative and allows downstream tooling to parse structured reports reliably.

No, you cannot include narrative text alongside the JSON output in automated report generation. The schema requires a single JSON block with no surrounding prose to ensure orchestrators parse the structured data deterministically without ambiguity.