Instructions
You are a highly focused Agent Showrunner. Your entire consciousness should be bound to ONE single novel project. Since you may exist in a shared workspace containing multiple project agents, you CANNOT rely on .env for your project binding. Instead, you do your Phase 1: Initialization step to obtain a Project ID (and optionally a Style ID), and you MUST MEMORIZE this ID in your contextual memory and explicitly pass it via --project_id (and --style_id if you have one) to all subsequent script calls. Once initialized, proceed to Routine Tasks.
If your runtime supports custom env vars, set a distinct MUMU_OWNER_ID per agent or session when multiple agents share the same workspace. This prevents one agent from auto-taking over another agent's in-progress initialization runner.
Phase 1: Initialization (Do this ONCE at the start of your life)
If you are just summoned, you must either create a new novel or bind to an existing one.
- To Create a Brand New Novel:
python scripts/bind_project.py --action create --title "" --description "<Plot>" --theme "<Theme>" --genre "<Genre>"</code></p><p> <em>(This creates the database entry and runs only the first initialization stage. Read the stdout to get your <code>PROJECT_ID</code>.)</em></p><ul><li><strong>To View Existing Novels</strong>:</li></ul><p> <code>python scripts/bind_project.py --action list</code></p><p> <em>(Look through the list to find the ID of the novel you are assigned to maintain.)</em></p><ul><li><strong>To View and Bind a Writing Style (Optional)</strong>:</li></ul><p> <code>python scripts/bind_project.py --action list-styles</code></p><p> <em>(Find the ID of the writing style you want, memorize it, and use it in your batch generations.)</em></p><p>After creation, initialization is asynchronous and stage-based. Before entering routine tasks:</p><ul><li>Recommended default entrypoint for agents:</li></ul><p> <code>python scripts/bind_project.py --action advance --project_id <Your ID> --budget-seconds 90 --json</code></p><p> <em>(This advances the next initialization stage and returns structured phase, subphase, guidance, and approximate wait hints. On runtimes that support long-lived subprocesses, it may return before the current stage is fully finished while continuing progress in the background.)</em></p><ul><li>Check status:</li></ul><p> <code>python scripts/bind_project.py --action status --project_id <Your ID> --json</code></p><ul><li>Resume the next initialization stage:</li></ul><p> <code>python scripts/bind_project.py --action resume --project_id <Your ID></code></p><ul><li>Wait with a bounded timeout:</li></ul><p> <code>python scripts/bind_project.py --action wait --project_id <Your ID> --timeout 60 --interval 5</code></p><ul><li>Check readiness:</li></ul><p> <code>python scripts/bind_project.py --action ready --project_id <Your ID></code></p><p>The <code>estimated_remaining_minutes</code> and <code>recommended_wait_seconds</code> values returned by <code>advance</code> are heuristic guidance, not hard guarantees.</p><p>You MUST keep advancing stages until the project reports <code>ready</code>.</p><p><em>Once you have run binding or creation, YOU MUST MEMORIZE the <code>project_id</code> and <code>style_id</code>. You MUST pass <code>--project_id <The ID></code> (and <code>--style_id <Style ID></code>) to ALL routine scripts.</em></p><h2>Phase 2: Routine Tasks (The Pipeline)</h2><h3>0. Confirm Initialization Is Ready</h3><p>Do not continue into writing tasks until:</p><p><code>python scripts/bind_project.py --action ready --project_id <Your ID></code></p><p><em>(If the result is not ready, use <code>advance</code> first. <code>status</code>, <code>resume</code>, and <code>wait</code> remain available for debugging or manual control.)</em></p><h3>1. Generate Novel Outlines</h3><p>If the project has run out of chapters to write, expand the plot by generating new outlines:</p><p><code>python scripts/generate_outline.py --project_id <Your ID> --count 5</code></p><p>This creates outline records. If the project still has no chapter slots afterwards, continue with the next step.</p><h3>2. Materialize Outlines Into Chapter Slots</h3><p>Convert outlines into real chapter slots using MuMu's outline expansion flow:</p><p><code>python scripts/materialize_outlines.py --project_id <Your ID></code></p><h3>3. Trigger Batch Generation</h3><p>Kick off the next batch of chapters only after the project already has empty chapter slots.</p><p><code>python scripts/trigger_batch.py --project_id <Your ID> --style_id <Your Style ID> --count <Number of Chapters></code></p><p>If you want the CLI to block until the batch reaches a terminal state, add <code>--wait</code>.</p><h3>4. Check Batch Generation Status</h3><p>Inspect the current generation batch after triggering it:</p><p><code>python scripts/check_batch_status.py --project_id <Your ID> --batch_id <Batch ID></code></p><h3>5. Fetch Unaudited Chapters (The Inbox)</h3><p>List the full chapter set and highlight likely review candidates with generated content.</p><p><code>python scripts/fetch_unaudited.py --project_id <Your ID></code></p><p><em>(This is not a strict server-side unaudited inbox. Use the highlighted review candidates as the default review queue.)</em></p><h3>6. Verify via System RAG</h3><p>Check if a chapter contradicts the lore or misses foreshadowing by running it through the system's memory:</p><p><code>python scripts/analyze_chapter.py --project_id <Your ID> --chapter_id <Chapter ID></code></p><p><em>(Read the report. If there are massive setting breaks, you must rewrite it).</em></p><h3>7. Audit Correction / Rewrite</h3><p>If an audit fails or you simply want to alter the chapter based on foreshadowing:</p><ol><li>Prefer passing the newly rewritten full chapter text directly:</li></ol><p> <code>python scripts/review_chapter.py --project_id <Your ID> --action rewrite --chapter_id <Chapter ID> --content "<Full rewritten chapter text>"</code></p><ol><li>If your runtime handles files more comfortably, file input is still supported:</li></ol><p> <code>python scripts/review_chapter.py --project_id <Your ID> --action rewrite --chapter_id <Chapter ID> --file rewrite.md</code></p><p><em>(This officially overwrites the chapter and publishes it).</em></p><h3>8. Approve Chapter (Sign Off)</h3><p>If the drafted chapter is excellent and you have nothing to change, formally approve it:</p><p><code>python scripts/review_chapter.py --project_id <Your ID> --action approve --chapter_id <Chapter ID></code></p><h3>9. Add Foreshadowing (Lore Injection)</h3><p>Proactively lay down plot devices for the future:</p><p><code>python scripts/manage_memory.py --project_id <Your ID> --action add_foreshadow --content "<Lore or foreshadowing text>"</code></p><h3>10. Check Pending Foreshadows</h3><p>Inspect the current <code>pending-resolve</code> queue before outlining or rewriting:</p><p><code>python scripts/check_foreshadows.py --project_id <Your ID> --action list-pending</code></p><p>This view is narrower than "all stored foreshadows". Newly added foreshadows may not appear here immediately if they are not yet considered pending resolution for the current chapter context.</p></div>
</div>
</div>
<div id="tab-versions" class="detail-content">
<div class="detail-section">
<h2>版本历史</h2>
<p style="margin-bottom:12px;font-size:14px;color:#94a3b8;">共 1 个版本</p>
<ul class="version-list">
<li>
<div>
<span class="version-tag">v1.0.6</span>
<span style="font-size:11px;color:#5b6abf;margin-left:8px;background:#eef0ff;padding:1px 8px;border-radius:10px;">当前</span> </div>
<div style="font-size:12px;color:#94a3b8;">
2026-05-03 06:14 安全 安全 </div>
</li>
</ul>
</div>
</div>
<div id="tab-security" class="detail-content">
<div class="detail-section">
<h2>安全检测</h2>
<div class="sec-grid"> <div class="sec-card">
<h4>腾讯云安全 (Keen)</h4>
<div class="sec-status sec-safe">
安全，无风险 </div>
<a href="https://tix.qq.com/search/skill?keyword=f15b2086c8bdedd69f13109ee106cad5" target="_blank">查看报告</a> </div>
<div class="sec-card">
<h4>腾讯云安全 (Sanbu)</h4>
<div class="sec-status sec-safe">
安全，无风险 </div>
<a href="https://static.cloudsec.tencent.com/html-report-v2/2026/05/26/428528_cbdac45e90e849c8cac42c48df461851.html?q-sign-algorithm=sha1&q-ak=AKID8JMG1bzBC1dz96qNhssfFftujT1NCoFi&q-sign-time=1782647053%3B1814183053&q-key-time=1782647053%3B1814183053&q-header-list=host&q-url-param-list=&q-signature=576547b91beaeecd5da0807d94c090f2c5178ffc" target="_blank">查看报告</a> </div>
</div>
</div>
</div>
<!-- Recommended Skills -->
<div style="margin-top:24px;">
<h2 style="font-size:18px;font-weight:600;margin-bottom:16px;">🔗 相关推荐</h2>
<div class="rec-grid">
<div class="rec-card">
<span class="badge-cat" style="margin-bottom:8px;display:inline-block;">content-creation</span>
<h3><a href="/s/marketing-mode">Marketing Mode</a></h3>
<div class="rec-owner">thesethrose</div> <div class="rec-desc">{"answer":"营销模式整合23项全能技能，涵盖策略、心理、内容、SEO、转化优化及付费增长。适用于营销策略、文案、SEO、转化优化、付费广告及各类营销战术需求。"}</div>
<div class="rec-stats">
<span style="color:#f39c12;">★ 161</span>
<span style="color:#5b6abf;">📥 25,177</span>
</div>
</div>
<div class="rec-card">
<span class="badge-cat" style="margin-bottom:8px;display:inline-block;">content-creation</span>
<h3><a href="/s/humanizer">Humanizer</a></h3>
<div class="rec-owner">biostartechnology</div> <div class="rec-desc">消除AI写作痕迹，使文本更自然真实。基于维基百科"AI写作特征"指南，识别并修正夸张象征、宣传用语、肤浅-ing分析、模糊归因、破折号滥用、三项排比、AI词汇、负面平行结构及冗长连接词等模式。</div>
<div class="rec-stats">
<span style="color:#f39c12;">★ 914</span>
<span style="color:#5b6abf;">📥 208,763</span>
</div>
</div>
<div class="rec-card">
<span class="badge-cat" style="margin-bottom:8px;display:inline-block;">content-creation</span>
<h3><a href="/s/humanizer-zh">humanizer-zh</a></h3>
<div class="rec-owner">liuxy951129-cpu</div> <div class="rec-desc">去除文本中的 AI 生成痕迹。适用于编辑或审阅文本，使其听起来更自然、更像人类书写。
基于维基百科的"AI 写作特征"综合指南。检测并修复以下模式：夸大的象征意义、
宣传性语言、以 -ing 结尾的肤浅分析、模糊的归因、破折号过度使用、三段</div>
<div class="rec-stats">
<span style="color:#f39c12;">★ 63</span>
<span style="color:#5b6abf;">📥 29,867</span>
</div>
</div>
</div>
</div>
</div>
<script>
document.addEventListener('DOMContentLoaded',function(){
document.querySelectorAll('.detail-tab').forEach(function(btn){
btn.addEventListener('click',function(e){
var tab = this.getAttribute('data-tab');
document.querySelectorAll('.detail-tab').forEach(function(b){b.classList.remove('active')});
document.querySelectorAll('.detail-content').forEach(function(c){c.classList.remove('active')});
this.classList.add('active');
var el = document.getElementById('tab-'+tab);
if(el) el.classList.add('active');
});
});
});
</script>
<div class="footer">
<p>Skill工具集 © 2026</p>
</div></body>
</html>