// Use when predicting Threads post performance, analyzing post history patterns, estimating engagement ceiling for a draft, or deciding what content type to write next. Works with or without personal data — uses darkseoking benchmark as fallback.
| name | darkseoking-post-predictor |
| description | Use when predicting Threads post performance, analyzing post history patterns, estimating engagement ceiling for a draft, or deciding what content type to write next. Works with or without personal data — uses darkseoking benchmark as fallback. |
Analyzes historical Threads post data using algorithm knowledge from darkseoking-mindset to predict post performance and recommend optimal content strategy.
references/personal-profile.md for future useMinimum useful data per post: content (or topic summary) + at least one engagement metric (likes, comments, or reposts).
For V2 dual-stage prediction (strongly recommended): views + likes per post. Without views, prediction falls back to V1 single-stage and cannot distinguish distribution-driven vs conversion-driven success.
Ideal CSV columns: content, likes, views, replies, reposts, shares, engagement_rate, media_type, is_quote, created_at
Minimum posts: 15+ for meaningful baseline. 30+ for pattern extraction. Under 15 — use darkseoking benchmark with caveats.
Accepted formats: CSV, pasted list, verbal description of recent 5-10 posts with approximate engagement numbers.
Personal profile: If references/personal-profile.md exists, load it to skip re-running full historical analysis. If it doesn't exist and user provides CSV, run historical-analysis.md and save the output as references/personal-profile.md. Profiles should be refreshed when new data is available (e.g. monthly).
| Scenario | Action |
|---|---|
| User provides full post history | Run complete analysis + prediction |
| User asks "what should I write next?" with data context | Run content-type recommendation |
| User wants to know ceiling before posting | Run prediction on draft + historical context |
| User wants to understand why a post underperformed | Run gap analysis against historical patterns |
| User has no data, just wants general guidance | Use darkseoking benchmark patterns from prediction-model.md |
darkseoking-mindset/references/darkseoking-all-posts.csvdarkseoking-mindset skillUse when writing, reviewing, or deciding what to post on Threads. Also triggers when planning post-viral strategy or checking posting timing. Runs pre-publish checks based on Meta algorithm patents.
Claude Code 專案配置審計。觸發：review/優化 CLAUDE.md、skills、settings、定期清洗累積內容、新專案上線前檢查。
Agent 配置設計指南 — 基於 Claude Code 6 個 built-in agent 的逆向分析。Use when: 設計新 agent、優化現有 agent prompt、決定工具/模型配置、撰寫 dispatch prompt。
AI Agent 成本工程 — 基於 Claude Code 的成本追蹤、prompt cache 最佳化、token 預算控制逆向分析。Use when: 優化 token 消耗、設計成本控制機制、分析 cache 效率、選擇模型配置。
Harness Engineering 設計模式 — 基於 Claude Code 原始碼逆向分析的 12 條可遷移原則。Use when: 設計 agent 系統架構、實作 tool orchestration、設計 context 管理策略、建構 agent loop。
System Prompt 工程 — 基於 Claude Code 914 行系統提示詞的逆向分析。Use when: 撰寫 system prompt、設計 prompt 動態組裝、最佳化 prompt cache 效率、撰寫安全指令。