Skip to content

notque/vexjoy-agent / skills/meta/do

Classify user requests and route to the correct skill, pipeline, and stack. Primary entry point for all delegated work.

references/perspective-prompts.md

View on GitHub
# Perspective Prompt Templates

Templates for each of the 10 parallel analysis agents. Each agent receives the full source material but analyzes through ONE lens only, ensuring true independence.

---

## Prompt Structure

Every perspective prompt follows this pattern:

```
Analyze the following source material focusing ONLY on [PERSPECTIVE]:

SOURCE MATERIAL:
{source_material}

YOUR PERSPECTIVE: [Perspective Name]
- [Focus area 1]
- [Focus area 2]
- [Focus area 3]
- [Focus area 4]

TASK:
1. Identify 5-7 key patterns
2. Extract 3-5 actionable rules
3. Explain application to {target_name}
4. Provide 2-3 concrete examples from source

Analyze only the assigned perspective so each agent produces an independent view.

Output format:
## [Perspective Name] Analysis

**Key Patterns**:
1. [Pattern with example]
...

**Extracted Rules**:
1. [Actionable rule]
...

**Application to {target_name}**:
[How patterns should influence the target]
```

---

## The 10 Perspectives

### 1. Structural Analysis
- Document organization and hierarchy
- Flow and transitions between sections
- Information architecture patterns
- How complexity is scaffolded

### 2. Clarity and Precision
- Language clarity techniques
- Precision in technical explanations
- Ambiguity elimination strategies
- Definition and term usage patterns

### 3. Technical Explanation Patterns
- How technical concepts are introduced
- Simplification without oversimplification
- Analogy and metaphor usage
- Layered explanation techniques

### 4. Audience Assumption Patterns
- What knowledge is assumed vs. explained
- How prerequisites are handled
- Jargon usage and introduction
- Progressive disclosure of complexity

### 5. Evidence and Citation Strategy
- How claims are supported
- Use of examples and concrete data
- Citation and reference patterns
- Authority establishment techniques

### 6. Narrative Progression
- Story arc and momentum
- Reader engagement techniques
- Hook and payoff patterns
- Tension and resolution structures

### 7. Paragraph and Sentence Architecture
- Sentence length and variety
- Paragraph structure and length
- Topic sentence usage
- Rhythm and cadence patterns

### 8. Header and Signposting Strategy
- Section naming conventions
- Preview and summary usage
- Reader orientation techniques
- Navigation and wayfinding patterns

### 9. Complexity Management
- How difficult topics are approached
- Gradual complexity increase
- Simplification strategies
- When to be thorough vs. concise

### 10. Limitation and Nuance Handling
- How caveats are presented
- Edge case discussion
- Uncertainty acknowledgment
- Trade-off presentation patterns

---

## Implementation Notes

- All 10 Task invocations MUST be in a single message for true parallel execution
- Each agent receives the FULL source material
- Each agent receives ONLY their assigned perspective
- Agents have zero knowledge of other agents' work (true independence)
- Each agent should produce 200-500 words of focused analysis
- Wait for ALL agents to complete before proceeding to synthesis

---

## Synthesis Template

After all 10 agents return results, create this synthesis document:

```markdown
# Synthesized Improvement Recommendations

## Common Themes Across Perspectives
[Patterns that appeared in 4+ perspectives]

## Unique Insights
[Valuable patterns from individual perspectives]

## Priority Rules for [Target]

### Must-Have (Priority 1)
1. [Rule - present in 7+ perspectives or critical impact]

### Should-Have (Priority 2)
1. [Rule - present in 4-6 perspectives or high impact]

### Nice-to-Have (Priority 3)
1. [Rule - present in 1-3 perspectives or moderate impact]

## Implementation Guidance
[Specific sections to add/modify in target]
[Concrete examples and templates]
```

---

## Completion Report Template

```markdown
# Parallel Multi-Perspective Analysis - Completion Report

**Target**: {target_name}
**Source Material**: {source_path}
**Token Budget Used**: ~{estimated_tokens} tokens

## Phase 1: Validation

- Target: [confirmed]
- Source: [confirmed, word count]

## Phase 2: Multi-Perspective Analysis

Successfully analyzed from 10 perspectives:
1. Structural Analysis - [Key insight]
2. Clarity and Precision - [Key insight]
3. Technical Explanation Patterns - [Key insight]
4. Audience Assumption Patterns - [Key insight]
5. Evidence and Citation Strategy - [Key insight]
6. Narrative Progression - [Key insight]
7. Paragraph and Sentence Architecture - [Key insight]
8. Header and Signposting Strategy - [Key insight]
9. Complexity Management - [Key insight]
10. Limitation and Nuance Handling - [Key insight]

**Total Rules Extracted**: {count}

## Phase 3: Synthesis

**Common Themes**: {count} patterns
**Priority 1 (Must-Have)**: {count} rules
**Priority 2 (Should-Have)**: {count} rules
**Priority 3 (Nice-to-Have)**: {count} rules

## Phase 4: Application

**Target Updated**: {target_name}
**Lines Added**: +{count}
**New Sections**: {count}

### Key Improvements:
1. [Improvement - contributing perspectives]
2. [Improvement - contributing perspectives]
3. [Improvement - contributing perspectives]

## Phase 5: Verification

**Depth Increase**: {before} lines -> {after} lines (+{percent}%)
**New Capabilities**: [List new patterns/rules]
**Integration Quality**: [Assessment]
```

---

## Good Source Material

### Excellent Sources
- Jakob Nielsen's usability articles (clear principles, concrete examples)
- Martin Fowler's refactoring patterns (systematic explanations)
- Stripe API documentation (progressive complexity, clear examples)
- Paul Graham's essays (narrative structure, persuasive techniques)
- PostgreSQL optimization guides (technical depth, practical patterns)

### Sources to Skip
- Marketing fluff without technical depth
- Auto-generated documentation
- Listicles without underlying methodology
- Opinion pieces without supporting evidence
- Code dumps without explanation
writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 322kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall