sop-code-review

---

name: sop-code-review description: Standardize code review execution with repeatable steps, evidence requirements, and confidence ceilings. allowed-tools:

• Read
• Write
• Edit
• Bash
• Glob
• Grep
• Task
• TodoWrite model: sonnet x-version: 3.2.0 x-category: quality x-vcl-compliance: v3.1.1 x-cognitive-frames:
• HON
• MOR
• COM
• CLS
• EVD
• ASP
• SPC

---

STANDARD OPERATING PROCEDURE

Purpose

Provide a concise, repeatable SOP for conducting code reviews that meet evidential and structural standards across repositories.

Trigger Conditions

• Positive: teams requesting a reference SOP, onboarding reviewers, or normalizing review quality across projects.
• Negative: deep-dive audits (use code-review-assistant) or runtime debugging (use functionality-audit).

Guardrails

• Confidence ceiling: Include Confidence: X.XX (ceiling: TYPE Y.YY) with ceilings {inference/report 0.70, research 0.85, observation/definition 0.95}.
• Evidence requirements: Every finding must include file:line, severity, and a reference (standard, guide, or metric).
• Structure-first: Ensure examples/tests show correct review phrasing, severity, and confidence statements.
• Adversarial validation: Spot-check for false positives and confirm blockers truly block functionality or safety.

Execution Phases

1. Preparation
   • Understand change intent, risk areas, and exclusions.
   • Set severity definitions and merge criteria (blocker/major/minor/nit).
2. Review Passes
   • Security and correctness first; then tests and completeness; finish with style/documentation.
   • Capture findings with evidence type and suggested fix.
3. Summary & Decision
   • Group findings, flag blockers, and propose next steps.
   • Provide approval recommendation and confidence with ceiling.
4. Follow-through
   • Track remediation owners and due dates; rerun spot checks after fixes.

Output Format

• Scope, exclusions, and severity definitions.
• Findings list with file:line, severity, evidence, and fix guidance.
• Approval recommendation and confidence statement.

Validation Checklist

• Scope and severity model documented.
• Findings include file:line, evidence, and references.
• Blockers vs. advisory items separated.
• Confidence ceiling provided; English-only output.

Confidence: 0.70 (ceiling: inference 0.70) - SOP rewritten to reflect Prompt Architect confidence discipline and Skill Forge structure-first review steps.