Deep Review reference

Multi-pass deep code review: security, quality, and test gap analysis with prioritized findings.

Classes

Class	Description
`DeepReviewAgentSDKWorkflow`	Orchestrates a multi-pass deep code review using Claude Agent SDK subagents: `security-reviewer`, `quality-reviewer`, and `test-gap-reviewer`.

Method	Parameters	Returns	Description
`execute`	`**kwargs: Any`	`WorkflowResult`	Runs the three specialized subagents and synthesizes their findings into a consolidated review report.

The workflow coordinates three subagents. Each reports findings independently; the orchestrator then synthesizes results into a single report.

Name	Domain
`security-reviewer`	Security vulnerabilities and risk findings
`quality-reviewer`	Code quality findings
`test-gap-reviewer`	Test coverage gaps

The synthesized output contains the following sections:

Section	Content
`## Summary`	Overall code health score (0–100), a 2–3 sentence executive summary, and finding counts by severity.
`## Security`	Findings from the security reviewer, ordered by severity.
`## Quality`	Findings from the quality reviewer, ordered by severity.
`## Test Gaps`	Findings from the test gap reviewer, ordered by priority.
`## Suggestions`	Top 5–10 actionable next steps ordered by impact, each referencing the specific finding it addresses.