notque/vexjoy-agent / skills/process/workflow
workflow
VerifiedStructured work: multi-phase tasks, feature builds, planning, objective loops, hill climbing.
references/comprehensive-review/references/output-templates.md
View on GitHub# Output Templates
## Findings Summary Matrix
Used during Wave 0+1+2 aggregate (Phase 3b) and final aggregate (Phase 3d):
```
| Agent | Wave | CRITICAL | HIGH | MEDIUM | LOW |
|--------------------------|------|----------|------|--------|-----|
| Per-Package: [pkg1] | 0 | N | N | N | N |
| Per-Package: [pkg2] | 0 | N | N | N | N |
| Per-Package: [...] | 0 | N | N | N | N |
| **Wave 0 Subtotal** | **0**| **N** | **N**| **N** | **N**|
| Security | 1 | N | N | N | N |
| Business Logic | 1 | N | N | N | N |
| Architecture | 1 | N | N | N | N |
| Silent Failures | 1 | N | N | N | N |
| Test Coverage | 1 | N | N | N | N |
| Type Design | 1 | N | N | N | N |
| Code Quality | 1 | N | N | N | N |
| Comment Analyzer | 1 | N | N | N | N |
| Language Specialist | 1 | N | N | N | N |
| Docs & Config | 1 | N | N | N | N |
| ADR Compliance | 1 | N | N | N | N |
| Newcomer | 1 | N | N | N | N |
| **Wave 1 Subtotal** | **1**| **N** | **N**| **N** | **N**|
| Performance | 2 | N | N | N | N |
| Concurrency | 2 | N | N | N | N |
| API Contract | 2 | N | N | N | N |
| Dependency Audit | 2 | N | N | N | N |
| Error Messages | 2 | N | N | N | N |
| Dead Code | 2 | N | N | N | N |
| Naming Consistency | 2 | N | N | N | N |
| Observability | 2 | N | N | N | N |
| Config Safety | 2 | N | N | N | N |
| Migration Safety | 2 | N | N | N | N |
| **Wave 2 Subtotal** | **2**| **N** | **N**| **N** | **N**|
| Contrarian | 3 | — challenges — | — | — | — |
| Skeptical Senior | 3 | — challenges — | — | — | — |
| User Advocate | 3 | — challenges — | — | — | — |
| Meta-Process | 3 | — challenges — | — | — | — |
| SAPCC Structural | 3 | — challenges — | — | — | — |
| **Wave 3 Summary** | **3**| **N agreed** | **N challenged** | **N downgraded** | **N dismissed** |
| **TOTAL (post-Wave 3)** | | **N** | **N**| **N** | **N**|
```
Wave Agreement Summary (append after matrix):
```
| Agreement Level | Count | Action |
|-----------------|-------|--------|
| UNANIMOUS | N | Fix immediately — high confidence |
| MAJORITY | N | Fix, note challenge in report |
| CONTESTED | N | Needs human judgment — present both sides |
```
## Final Report Template
Write to `comprehensive-review-report.md`:
```markdown
# Comprehensive Code Review Report v4
**Date**: [date]
**Files reviewed**: [N]
**Packages discovered**: [N]
**Agents dispatched**: [N] (Wave 0: [N per-package], Wave 1: 12, Wave 2: 10, Wave 3: [4-5 adversarial])
**Total findings**: [N]
**Findings fixed**: [N]
**Findings blocked**: [N] (ONLY if fix breaks tests after alternative attempt — must be <10%)
---
## Verdict: [CLEAN | ALL_FIXED | BLOCKED_ITEMS]
- CLEAN: No findings (rare)
- ALL_FIXED: Every finding was fixed (expected outcome)
- BLOCKED_ITEMS: Some fixes break tests even after alternative attempts (<10%)
[2-3 sentences: Overall assessment. What systemic patterns emerged?
Is the codebase better after fixes?]
## Wave Summary
| Wave | Agents | Findings | Fixed | Unique to Wave |
|------|--------|----------|-------|----------------|
| Wave 0 (Per-Package) | N | N | N | N |
| Wave 1 (Foundation) | 12 | N | N | N |
| Wave 2 (Deep-Dive) | 10 | N | N | N |
| Wave 3 (Adversarial) | N | — | — | N challenges |
| **TOTAL** | **N** | **N** | **N** | |
## Wave Agreement Analysis
| Agreement Level | Count | Fixed | Skipped | Human Decided |
|-----------------|-------|-------|---------|---------------|
| UNANIMOUS | N | N | — | — |
| MAJORITY | N | N | — | — |
| CONTESTED | N | N | N | N |
## Wave 0: Per-Package Results
| Package | Health | Files | Findings | Key Issue |
|---------|--------|-------|----------|-----------|
| [pkg/path1] | HEALTHY | N | N | — |
| [pkg/path2] | NEEDS_ATTENTION | N | N | [biggest] |
| ... | ... | ... | ... | ... |
**Cross-Package Patterns**: [List of patterns seen across multiple packages]
## Agent Summary
| Agent | Wave | Findings | Fixed | Blocked | Key Issue |
|-------|------|----------|-------|---------|-----------|
| Per-Package (total) | 0 | N | N | N | [biggest] |
| Security | 1 | N | N | N | [biggest] |
| Business Logic | 1 | N | N | N | [biggest] |
| Architecture | 1 | N | N | N | [biggest] |
| Silent Failures | 1 | N | N | N | [biggest] |
| Test Coverage | 1 | N | N | N | [biggest] |
| Type Design | 1 | N | N | N | [biggest] |
| Code Quality | 1 | N | N | N | [biggest] |
| Comment Analyzer | 1 | N | N | N | [biggest] |
| Language Specialist | 1 | N | N | N | [biggest] |
| Docs & Config | 1 | N | N | N | [biggest] |
| ADR Compliance | 1 | N | N | N | [biggest] |
| Newcomer | 1 | N | N | N | [biggest] |
| Performance | 2 | N | N | N | [biggest] |
| Concurrency | 2 | N | N | N | [biggest] |
| API Contract | 2 | N | N | N | [biggest] |
| Dependency Audit | 2 | N | N | N | [biggest] |
| Error Messages | 2 | N | N | N | [biggest] |
| Dead Code | 2 | N | N | N | [biggest] |
| Naming Consistency | 2 | N | N | N | [biggest] |
| Observability | 2 | N | N | N | [biggest] |
| Config Safety | 2 | N | N | N | [biggest] |
| Migration Safety | 2 | N | N | N | [biggest] |
| Contrarian | 3 | — | — | — | [key challenge] |
| Skeptical Senior | 3 | — | — | — | [key challenge] |
| User Advocate | 3 | — | — | — | [key challenge] |
| Meta-Process | 3 | — | — | — | [key challenge] |
| SAPCC Structural | 3 | — | — | — | [key challenge or N/A] |
| **TOTAL** | | **N** | **N** | **N** | |
## Context Cascade Effectiveness
How each wave's context helped later waves find deeper issues:
| Wave 2 Agent | Wave 0 Context Used | Wave 1 Context Used | Additional Findings Due to Context |
|-------------|--------------------|--------------------|-------------------------------------|
| Performance | Package complexity hotspots | Architecture hot paths | [N findings] |
| Concurrency | Intra-package concurrent patterns | Silent failures + arch | [N findings] |
| ... | ... | ... | ... |
### Wave 3 Challenge Effectiveness
How adversarial review changed the final outcome:
| Wave 3 Agent | Findings Challenged | Downgraded | Dismissed | Key Insight |
|-------------|--------------------|-----------|-----------|----|
| Contrarian | N | N | N | [biggest challenge] |
| Skeptical Senior | N | N | N | [biggest challenge] |
| User Advocate | N | N | N | [biggest challenge] |
| Meta-Process | N | N | N | [biggest challenge] |
| SAPCC Structural | N or N/A | N | N | [biggest challenge or N/A] |
## Findings by Severity
### CRITICAL
[Each finding with before/after code]
### HIGH
[Each finding with before/after code]
### MEDIUM
[Summary with file references]
### LOW
[Brief list]
## Contested Findings (Wave 3 vs Wave 1+2)
| Finding | Wave 1+2 Severity | Wave 3 Verdict | Resolution |
|---------|-------------------|----------------|------------|
| [summary] | HIGH | CHALLENGE: [reason] | [Fixed / Skipped / Human decided] |
| ... | ... | ... | ... |
## Quick Wins Applied
[List of easy fixes that improved quality]
## Blocked Items (if any — must be <10% of total)
[List of findings where fix AND alternative fix both break tests]
[Each must include: what was tried, why it failed, suggested manual approach]
## What's Done Well
[Genuine positives found during review]
## Systemic Recommendations
[2-3 big-picture patterns observed across findings]
```
## task_plan.md Template
Created at the start of Phase 1:
```markdown
# Task Plan: Comprehensive Review v4
## Goal
Four-wave review and auto-fix of [N] changed files across [N] packages.
## Phases
- [ ] Phase 0.5: Static Analysis (linters, auto-fix trivial, capture remaining)
- [ ] Phase 1: Scope (identify files, detect org, discover packages, create findings dir)
- [ ] Phase 1b: Wave 0 Dispatch (per-package deep review)
- [ ] Phase 1c: Wave 0 Aggregate (per-package findings)
- [ ] Phase 1.5: Library Contract Verification (Go repos only)
- [ ] Phase 2a: Wave 1 Dispatch (12 foundation agents + Wave 0 context)
- [ ] Phase 2b: Wave 1 Aggregate (collect and summarize Wave 0+1 findings)
- [ ] Phase 3a: Wave 2 Dispatch (10 deep-dive agents with Wave 0+1 context)
- [ ] Phase 3b: Wave 2 Aggregate (merge Wave 0+1+2 findings)
- [ ] Phase 3c: Wave 3 Dispatch (4-5 adversarial agents with Wave 0+1+2 context)
- [ ] Phase 3d: Wave 3 Aggregate (merge adversarial challenges, label agreement)
- [ ] Phase 4: Fix (auto-fix on branch)
- [ ] Phase 5: Report (write report, verify)
## Review Profile
- Files: [list]
- Packages discovered: [N]
- Wave 0 agents: [N] (one per package)
- Wave 1 agents: 12
- Wave 2 agents: 10
- Wave 3 agents: 4-5 (adversarial; 5 if SAPCC detected)
- Org conventions: [detected org or none]
- Mode: [review+fix | review-only]
## Findings Directory
$REVIEW_DIR = [path from Phase 1 Step 3]
## Status
**Currently in Phase 1** - Discovering packages
```
## Findings Directory Layout
```
$REVIEW_DIR/
wave0-findings.md — Per-package deep review results
wave1-findings.md — Foundation agent results (12 agents)
wave01-summary.md — Combined Wave 0+1 context for Wave 2
wave2-findings.md — Deep-dive agent results (10 agents)
wave012-summary.md — Combined Wave 0+1+2 context for Wave 3
wave3-findings.md — Adversarial challenge results (4-5 agents)
final-report.md — Aggregated, deduplicated, agreement-labeled
```
| File | Written By | Read By |
|------|-----------|---------|
| `$REVIEW_DIR/wave0-findings.md` | Phase 1c | Phase 2a, 2b |
| `$REVIEW_DIR/wave1-findings.md` | Phase 2b | Phase 3a |
| `$REVIEW_DIR/wave01-summary.md` | Phase 2b | Phase 3a |
| `$REVIEW_DIR/wave2-findings.md` | Phase 3b | Phase 3c |
| `$REVIEW_DIR/wave012-summary.md` | Phase 3b | Phase 3c |
| `$REVIEW_DIR/wave3-findings.md` | Phase 3d | Phase 4 |
| `$REVIEW_DIR/final-report.md` | Phase 3d | Phase 4, Phase 5 |
These files persist in `/tmp/` until next reboot and can be re-read in future sessions if needed.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies. | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |
| meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要". | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |