# Proactive Review Workflow

> Automate multi-pass code review across quality, security, and performance dimensions

- Skill: `diegosouzapw/proactive-review-workflow` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add diegosouzapw/proactive-review-workflow`
- Raw SKILL.md: https://api.skillmd.com/api/skills/diegosouzapw/proactive-review-workflow/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Security
- Author: diegosouzapw (https://skillmd.com/u/diegosouzapw)
- Updated: 2026-09-08
- Page: https://skillmd.com/skills/diegosouzapw/proactive-review-workflow

---


# Proactive Review Workflow Skill

Automate multi-pass code review across quality, security, and performance dimensions.

## Trigger Conditions
- PR opened or updated
- Merge event
- User invokes with "proactive review" or "full review"

## Input Contract
- **Required:** PR diff or file changes
- **Optional:** Review focus areas, prior review comments

## Output Contract
- Multi-dimensional review report (quality, security, performance)
- Categorized findings (blocking, suggestion, nit)
- Post-merge validation results

## Tool Permissions
- **Read:** PR diffs, source code, test files, configs
- **Write:** Review comments, review reports
- **Search:** Related code and patterns

## Execution Steps
1. Analyze PR diff for scope and affected components
2. Run quality pass (complexity, naming, error handling, dead code)
3. Run security pass (injection, auth, secrets, dependencies)
4. Run performance pass (N+1 queries, unbounded collections, missing indexes)
5. Categorize findings: blocking (must fix), suggestion (consider), nit (style)
6. Run post-merge validation on merged result
7. Generate review report

## Success Criteria
- All three dimensions reviewed (quality, security, performance)
- Findings categorized by severity
- No blocking findings remain unresolved
- Post-merge validation passes

## Escalation Rules
- Escalate if PR is >400 lines (diminishing review quality)
- Escalate if Critical security finding detected
- Escalate if architectural drift detected

## Example Invocations

**Input:** "Run proactive review on PR #42"

**Output:** Review: 3 files, 89 lines changed. Quality: 1 suggestion (extract helper function). Security: 1 blocking (SQL injection in search query — use parameterized query). Performance: 1 suggestion (add index for new WHERE clause). Post-merge: clean, no composition errors.