菜单
FanDuel NFL DFS slate prep for gambling-wiki W8 — read projection CSV exports (Stokastic/FantasyLabs), apply wiki GPP playbook, run pydfs lineup gen, write briefs/ slate card. Use when user says slate prep, FanDuel NFL, DFS lineups, Stokastic export, or Sunday NFL DFS.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use Cemini-adopted GEO/website tools (E-GEO rewrite prompts, geo-optimizer CLI, wondelai CRO skills) owned by the SEO/GEO wiki. Trigger from ANY Cursor workspace when rewriting local-business service pages for AI citation, auditing a site for GEO/AEO readiness, diagnosing why a brick-and-mortar site does not convert/book, or the user mentions E-GEO / geo-optimizer / wondelai CRO.
Multi-model Cursor audit — dispatches three parallel readonly subagents on different models to debug code, configs, plans, or briefs and synthesize consensus vs unique findings. Use when the user says cursor audit, /cursor-audit, multi-model audit, council audit, second-opinion debug, or wants independent model reviews before fixing or shipping.
Single-shot meta-prompt envelope for non-trivial tasks (/goal). Expands user intent into context, success criteria, operating rules, and completion proof. Use when the user invokes /goal or asks for a full-prompt goal session.
Super audit council — three Cursor readonly subagents plus two OpenRouter API legs (Fusion + premium) and one free local Ollama cross-check when available (6-model default). Falls back to 5-model (3 Cursor + 2 OpenRouter) without local. Operator may override routing. Use for super audit, /super-audit, council super audit, pre-ship multi-opinion review.
| name | nfl-fanduel-slate-prep |
| description | FanDuel NFL DFS slate prep for gambling-wiki W8 — read projection CSV exports (Stokastic/FantasyLabs), apply wiki GPP playbook, run pydfs lineup gen, write briefs/ slate card. Use when user says slate prep, FanDuel NFL, DFS lineups, Stokastic export, or Sunday NFL DFS. |
Operator stack: FanDuel GPP (not Underdog BBM7, not Hard Rock book). Read wiki playbook before building.
briefs/{season}-w{NN}-slate-hub.md first; do not re-research schedule/weather/injuryresearch to be indexed/ from Stokastic spreadsheet export or FantasyLabs model exportpython3 scripts/normalize_dfs_projection_csv.py \
--in "research to be indexed/<export>.csv" \
--out "research to be indexed/fanduel-nfl-pydfs-YYYY-MM-DD.csv" \
--source stokastic
If normalize fails, inspect headers — ensure Team column exists in paid export. Output must match FanDuel pydfs columns (see @entities/tools/pydfs-lineup-optimizer.md).
pip install pydfs-lineup-optimizer # once; MIT license — see tool page
python3 scripts/fanduel_slate_optimize.py \
--csv "research to be indexed/fanduel-nfl-pydfs-YYYY-MM-DD.csv" \
--count 150 \
--max-exposure 0.35 \
--stack qb:2 \
--out "briefs/fanduel-lineups-YYYY-MM-DD.csv"
Tune --count and --stack per slate size. FanDuel upload expects player names matching site IDs — verify a sample lineup in FD upload UI before mass entry.
briefs/YYYY-MM-DD_fanduel-nfl-slate.md)Include:
| Section | Content |
|---|---|
| Slate environment | Top 3 game stacks (Vegas totals, both QBs) |
| Chalk RBs | High ownership backs to accept |
| Leverage WR/TE | Low-owned pass-catchers with stack fit |
| Core locks | User/core plays in all or most lineups |
| Exposure notes | Max exposure settings used |
| Upload file | Path to briefs/fanduel-lineups-*.csv |
Apply K124 rules: 3 game environments × 2 QBs for MME; RB in flex on half-PPR; avoid TE flex; leave ≤$500 salary unused.
If user has Stokastic/Labs UI only: produce slate brief + stack plan from wiki + injury news; skip pydfs step until export available.
No MCP server exists for any major DFS optimizer. Integration = CSV drop + scripts above.