$npx -y skills add Soul-Brews-Studio/arra-oracle-skills-cli --skill oracle-prismMulti-perspective analysis — one agent transforms through 5 lenses (Archaeologist, Bug Hunter, Skeptic, Architect, Auditor) inline, no subagents. Use when user says "prism", "oracle-prism", "multi-perspective", "มองหลายมุม", "did we miss anything", "analyze from all angles", or b
| 1 | # /oracle-prism — Multi-Perspective Analysis |
| 2 | |
| 3 | > "แสงเดียวผ่านปริซึม แตกเป็นหลายสี — เรื่องเดียวกัน มองจากหลายมุม เห็นต่างกัน" |
| 4 | |
| 5 | One agent, five perspectives, no subagents. Transform inline — shift between lenses to dig through work, decisions, or problems from angles that a single viewpoint would miss. |
| 6 | |
| 7 | ## Usage |
| 8 | |
| 9 | ``` |
| 10 | /oracle-prism # Analyze current session (default) |
| 11 | /oracle-prism "the rename migration" # Analyze a specific topic |
| 12 | /oracle-prism --lenses 3 # Use 3 lenses instead of 5 |
| 13 | /oracle-prism --custom "Security Auditor,Performance Engineer,UX Designer" |
| 14 | ``` |
| 15 | |
| 16 | ## How It Works |
| 17 | |
| 18 | **No subagents.** The main agent transforms between perspectives sequentially, producing one section per lens. Each lens sees the same facts but asks different questions. |
| 19 | |
| 20 | This is NOT adversarial analysis (which tries to disprove). This is **prismatic analysis** — same light, different colors. |
| 21 | |
| 22 | ## Default 5 Lenses |
| 23 | |
| 24 | | # | Lens | Emoji | Question it asks | |
| 25 | |---|------|-------|-----------------| |
| 26 | | 1 | Archaeologist | `🔍` | "What actually happened? Timeline, sequence, facts." | |
| 27 | | 2 | Bug Hunter | `🐛` | "What problems did we hit? What's still broken?" | |
| 28 | | 3 | Skeptic | `💀` | "What did we do wrong? What would we redo?" | |
| 29 | | 4 | Architect | `🏗️` | "What changed structurally? What's the before/after?" | |
| 30 | | 5 | Auditor | `📋` | "What's left undone? What's inconsistent?" | |
| 31 | |
| 32 | ## Alternate Lens Sets |
| 33 | |
| 34 | Use `--custom` to define your own, or pick a preset: |
| 35 | |
| 36 | ### `--preset retro` (session retrospective) |
| 37 | | Lens | Question | |
| 38 | |------|----------| |
| 39 | | Historian | What happened, in what order? | |
| 40 | | Critic | What went poorly or slowly? | |
| 41 | | Cheerleader | What went well? What should we repeat? | |
| 42 | | Connector | What patterns connect to past work? | |
| 43 | | Planner | What's the next move? | |
| 44 | |
| 45 | ### `--preset design` (design review) |
| 46 | | Lens | Question | |
| 47 | |------|----------| |
| 48 | | User | Is this easy to use? What's confusing? | |
| 49 | | Maintainer | Is this easy to change later? | |
| 50 | | Breaker | How can this fail? Edge cases? | |
| 51 | | Simplifier | What can be removed? | |
| 52 | | Integrator | How does this fit with everything else? | |
| 53 | |
| 54 | ### `--preset incident` (post-incident) |
| 55 | | Lens | Question | |
| 56 | |------|----------| |
| 57 | | Firefighter | What happened and how was it fixed? | |
| 58 | | Detective | What was the root cause chain? | |
| 59 | | Defender | What guards existed? Why didn't they catch it? | |
| 60 | | Forecaster | What similar things could happen next? | |
| 61 | | Builder | What systemic fix prevents recurrence? | |
| 62 | |
| 63 | ## Output Format |
| 64 | |
| 65 | ```markdown |
| 66 | ## 🔍 Lens 1: Archaeologist — "What happened?" |
| 67 | |
| 68 | [Timeline table or narrative] |
| 69 | |
| 70 | --- |
| 71 | |
| 72 | ## 🐛 Lens 2: Bug Hunter — "What broke?" |
| 73 | |
| 74 | [Problems found, with evidence] |
| 75 | |
| 76 | --- |
| 77 | |
| 78 | ## 💀 Lens 3: Skeptic — "What went wrong?" |
| 79 | |
| 80 | [Mistakes, with what should have been done] |
| 81 | |
| 82 | --- |
| 83 | |
| 84 | ## 🏗️ Lens 4: Architect — "What changed?" |
| 85 | |
| 86 | [Before/after structural view] |
| 87 | |
| 88 | --- |
| 89 | |
| 90 | ## 📋 Lens 5: Auditor — "What's left?" |
| 91 | |
| 92 | [Table of pending items with status and urgency] |
| 93 | |
| 94 | --- |
| 95 | |
| 96 | **Cross-lens summary:** [2-3 sentences synthesizing what multiple lenses agree on] |
| 97 | ``` |
| 98 | |
| 99 | ## Rules |
| 100 | |
| 101 | 1. **No subagents** — all lenses run in the main agent, sequentially |
| 102 | 2. **Each lens gets a section** with its emoji, name, and guiding question as header |
| 103 | 3. **Evidence required** — cite files, commits, commands, timestamps. No vague claims |
| 104 | 4. **Lenses disagree** — if the Architect says "clean" but the Auditor says "incomplete", show both. Don't harmonize |
| 105 | 5. **Cross-lens summary at the end** — what do multiple lenses converge on? |
| 106 | 6. **Tables over prose** — use tables for timelines, status lists, before/after comparisons |
| 107 | 7. **Session context by default** — if no topic given, analyze the current session's work |
| 108 | 8. **3-7 lenses** — minimum 3, maximum 7. Default 5. Use `--lenses N` to adjust |
| 109 | |
| 110 | ## When to Use |
| 111 | |
| 112 | | Situation | Why prism helps | |
| 113 | |-----------|----------------| |
| 114 | | End of session | Catch what you missed before closing | |
| 115 | | After a migration/rename | Find inconsistencies across the change | |
| 116 | | Design decision | See tradeoffs from user/maintainer/breaker angles | |
| 117 | | Post-incident | Structured multi-angle without blame | |
| 118 | | "Did we miss anything?" | The Auditor lens exists for this | |
| 119 | |
| 120 | ## Relationship to Other Skills |
| 121 | |
| 122 | | Skill | Pattern | Agents | |
| 123 | |-------|---------|--------| |
| 124 | | `/oracle-prism` | Multi-perspective, same agent | 0 (inline transform) | |
| 125 | | `/adversarial-analysis` | Try to disprove a claim | 5 parallel subagents | |
| 126 | | `/rrr` | Session retrospective | 0 (or 5 in --deep) | |
| 127 | | `/roundtable` | Discussion between pe |