{"id":20646,"plugin_id":"plugins_6aa5c14ce164819185474f73c890b680","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T23:16:24.367Z","digest":"8f87babaf5058f93bd2c9d15ff0495fc47ec8f0dc721cd60248d2b1f0ba62b01","against":null,"payload":{"description":"Convene a panel of reviewers to analyze a draft from multiple perspectives, then synthesize their feedback into consensus findings, productive tensions, and prioritized recommendations.","included_files":[],"name":"cw-panel","skill_md_contents":"---\nname: cw-panel\ndescription: Convene a panel of reviewers to analyze a draft from multiple perspectives, then synthesize their feedback into consensus findings, productive tensions, and prioritized recommendations.\n---\n\n# Panel Review\n\nConvene multiple reviewer agents to analyze a piece, then synthesize their perspectives into a unified analysis that surfaces consensus, tensions, and priorities.\n\n## When to Use\n\n- Before publishing something high-stakes\n- When you want multiple perspectives without running reviewers manually\n- When you suspect different aspects need different kinds of attention\n- When you're too close to the work to know what feedback you need\n\n## The Flow\n\n### 1. Load Context\n\nRead `../../references/context-contract.md`, the draft, and any available brief, outline, project instructions, feedback, or status metadata to understand:\n- **piece_type** — essay, argument, explainer, narrative, etc.\n- **audience** — general, specialist, insider\n- **stage** — early draft, revision, near-final\n- **goals** — if stated\n\nInfer missing context from the draft and request. Do not require `.status.yaml`.\n\n### 2. Propose Panel\n\nBased on piece context, propose 4-6 reviewers.\n\n**Selection heuristics:**\n\n| Signal | Suggests Including |\n|--------|-------------------|\n| Personal/reflective content | sedaris, mom |\n| Argumentative content | nemesis, vonnegut |\n| Technical or jargon-heavy | reader, mom, hemingway |\n| Narrative structure | vonnegut, hitchcock, sorkin |\n| Feels slow or wandering | sorkin |\n| Feels bloated | hemingway |\n| High-stakes / pre-publish | nemesis, hemingway |\n| Cold read, uncertain audience response, or missing setup | reader |\n| General audience | reader |\n| Non-expert accessibility | mom |\n| Early stage | vonnegut (fundamentals) |\n| Late stage | hemingway, nemesis (polish, stress-test) |\n\n**Check the active writer and project context for:**\n- Preferred reviewers\n- Reviewers to exclude\n- Past patterns\n\n**Present the proposal:**\n\n```\n## Proposed Panel\n\n**Draft:** [title or slug]\n**Context:** [piece_type] for [audience], currently at [stage]\n\nBased on this context, I recommend:\n\n| Reviewer | Why |\n|----------|-----|\n| **[name]** | [One-line rationale tied to piece context] |\n| **[name]** | [Rationale] |\n| **[name]** | [Rationale] |\n| **[name]** | [Rationale] |\n\n**Not including:**\n- **[name]** — [Why not relevant for this piece]\n- **[name]** — [Why not]\n\nProceed with this panel, or adjust?\n```\n\nWait for user confirmation or modification.\n\n### 3. Run Reviewers in Parallel\n\nOnce confirmed:\n\n1. Run all selected reviewers in parallel when the active platform supports parallel workers and the user's explicit panel request authorizes them.\n   - On Claude, use the matching `compound-writing:review:*` subagent when available.\n   - On Codex or another runtime, use the matching reviewer skill or an isolated worker with that lens.\n   - Give each reviewer the draft plus the minimum relevant context.\n   - Each reviewer returns its standard prose output independently.\n2. Collect all outputs\n\n**Example spawning pattern:**\n\n```\n[Spawn in parallel:]\n- Task: nemesis-reviewer with draft\n- Task: mom-reviewer with draft\n- Task: hemingway-reviewer with draft\n- Task: sorkin-reviewer with draft\n```\n\n### 4. Run Synthesizer\n\nUse the panel-synthesizer agent when available; otherwise synthesize the outputs directly. Include:\n- The draft\n- Draft context from the active brief, instructions, metadata, or careful inference\n- All reviewer outputs (full prose)\n\n**Synthesizer prompt should include:**\n\n```\n## Draft\n\n[full draft text]\n\n## Context\n\npiece_type: [from context or inference]\naudience: [from context or inference]\nstage: [from context or inference]\ngoals: [from context if present]\n\n## Reviewer Outputs\n\n### nemesis-reviewer\n[full output]\n\n### mom-reviewer\n[full output]\n\n### hemingway-reviewer\n[full output]\n\n[etc.]\n```\n\n### 5. Return Synthesis\n\nPresent the synthesizer's output to the user.\n\n## Panel Size\n\n- **Minimum:** 3 reviewers (fewer than this, just run them individually)\n- **Recommended:** 4-5 reviewers (good coverage without redundancy)\n- **Maximum:** 8 reviewers (all of them—full stress test)\n\n## Available Reviewers\n\n| Reviewer | Subagent Type | Focus |\n|----------|---------------|-------|\n| nemesis | compound-writing:review:nemesis-reviewer | Logical rigor, unsupported claims |\n| reader | compound-writing:review:reader-reviewer | First-time reading experience, missing setup, trust and friction |\n| mom | compound-writing:review:mom-reviewer | Accessibility, jargon, general reader |\n| hemingway | compound-writing:review:hemingway-reviewer | Economy, cutting, word-level |\n| sedaris | compound-writing:review:sedaris-reviewer | Specificity, humor, voice |\n| sorkin | compound-writing:review:sorkin-reviewer | Pacing, momentum, forward motion |\n| vonnegut | compound-writing:review:vonnegut-reviewer | Story fundamentals, structure |\n| hitchcock | compound-writing:review:hitchcock-reviewer | Tension, suspense, stakes |\n\n## Default Panels by Piece Type\n\nIf the active context names a piece type, use it as a starting point:\n\n| piece_type | Default Panel |\n|------------|---------------|\n| `essay` / `personal` | reader, sedaris, vonnegut, sorkin |\n| `argument` / `opinion` | nemesis, vonnegut, hemingway, reader |\n| `explainer` / `technical` | reader, mom, hemingway, sorkin |\n| `narrative` / `story` | vonnegut, hitchcock, sorkin, sedaris |\n| `newsletter` | reader, sorkin, hemingway |\n\nAlways adjust based on specific signals in the draft.\n\n## Example Session\n\n```\nUser: /cw-panel\n\nAgent: I'll convene a panel to review your draft. Let me read it and check the status.\n\n[Reads draft]\n[Loads the brief and project context: personal essay, general audience,\n revision stage, goal: \"make it land emotionally without being sappy\"]\n\n## Proposed Panel\n\n**Draft:** father-and-the-sea\n**Context:** essay for general audience, revision stage\n\nBased on the context—personal essay aiming for emotional impact without\nsentimentality—I recommend:\n\n| Reviewer | Why |\n|----------|-----|\n| **mom** | General audience check. Will flag where you lose non-insiders. |\n| **sedaris** | Personal material that might benefit from more specificity or humor. |\n| **sorkin** | Revision stage—check if pacing is tight. |\n| **nemesis** | The emotional claims need to be earned, not asserted. |\n\n**Not including:**\n- **hemingway** — The prose reads clean; economy isn't the issue here\n- **hitchcock** — Not a tension-driven piece\n- **vonnegut** — Structure is solid; this needs polish not fundamentals\n\nProceed with this panel, or adjust?\n\nUser: Add hemingway actually, I think it might be overwritten\n\nAgent: Adding hemingway. Running the panel now.\n\n[Spawns 5 reviewer agents in parallel]\n[Collects outputs]\n[Spawns panel-synthesizer with draft + context + all outputs]\n\n## Panel Synthesis\n\n**Panel:** nemesis, hemingway, mom, sedaris, sorkin\n**Piece context:** essay for general audience, revision stage, goal: emotional\nimpact without sappiness\n\n---\n\n### Consensus Findings\n\n| Issue | Flagged By | Recommendation |\n|-------|------------|----------------|\n| Opening two paragraphs are slow | sorkin, hemingway, nemesis | Cut or restructure—start closer to the boat |\n| \"He taught me everything\" is unearned | nemesis, sedaris | Show what he taught, don't declare it |\n| Section 3 gets abstract | mom, sorkin | Ground it in a specific moment |\n\n---\n\n### Productive Tensions\n\n**Tension: The fishing gear paragraph**\n\n> \"His tackle box was a museum of failures—rusted lures from the '70s, a reel\n> that hadn't worked since Reagan, hooks that had never caught anything but\n> his own thumb.\"\n\n| Cut it | Keep it |\n|--------|---------|\n| **hemingway:** 47 words that could be 15. \"His tackle box held decades of failure\" does the job. | **sedaris:** This is the most alive moment in the piece. The specific details are doing real work. |\n\n**What's at stake:** The piece aims for emotional impact without sappiness.\nSpecific, wry details (sedaris) might be the path there. But if they slow\nmomentum, they work against the goal.\n\n---\n\n### Unique Insights\n\n- **sedaris:** \"There's a joke hiding in the part about him never admitting he\n  was wrong. You're circling it but won't commit.\"\n- **nemesis:** \"Your ending asks the reader to feel something you haven't set\n  up. The 'I finally understood' move is unearned.\"\n\n---\n\n### Recommended Priorities\n\n1. **Restructure the opening** — Three reviewers flagged this. Start on the\n   boat, not before it.\n2. **Earn the ending** — The emotional payoff needs setup. Show what you\n   finally understood through a scene, not a declaration.\n3. **Decide on the tackle box** — Genuine tension. Your call based on what\n   you're optimizing for.\n\n---\n\n### The Hard Question\n\n> What specifically did you learn from him that you couldn't have learned from\n> anyone else? Multiple reviewers noted you assert the relationship's importance\n> without showing what made it irreplaceable.\n```\n\n## Error Handling\n\n**If a reviewer fails to spawn:**\n- Note which reviewer is missing\n- Proceed with available outputs\n- Mention the gap in synthesis\n\n**If the draft has no metadata:**\n- Infer what you can from the draft and request.\n- State only assumptions that affect the panel or verdict.\n- Do not create a status file merely to run the review.\n\n**If user provides no draft:**\n- Ask for the draft path or slug to load\n- Or ask them to paste the draft\n\nFor tensions that deserve another round, offer `cw-debate`; do not automatically spend the extra time and context.\n\n## Lessons\n\n[Skill-specific lessons will be added here as they're captured]\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}