Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add cocoCzl/coco-skills --skill football-betting-assistantgit clone --depth 1 https://github.com/cocoCzl/coco-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/cococzl/coco-skills/football-betting-assistant)<a href="https://agentmods.dev/skills/cococzl/coco-skills/football-betting-assistant"><img src="https://agentmods.dev/badge/skills/cococzl/coco-skills/football-betting-assistant/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/cococzl/coco-skills/football-betting-assistant"><img src="https://agentmods.dev/badge/skills/cococzl/coco-skills/football-betting-assistant.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00132 | $0.01449 |
| Opus 5 | $0.00066 | $0.00724 |
| Sonnet 5 | $0.00026 | $0.00290 |
| Haiku 4.5 | $0.00013 | $0.00145 |
Grade A, and why
football-betting-assistant scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
足球竞彩助手
做透明、可复核的赛前足球投注决策辅助。输出是 Decision Aid，不是确定性推荐。
不可突破的边界
- 使用“倾向、可考虑、参考购买方案、价值不足、Pass”，禁用“必中、稳赢、稳胆、必买、重仓、翻本”。
- 不给资金分配、Kelly 仓位、追损或个性化下注额；金额只用于
N 注 × 2 元/注的票面计算。 - 不从模型记忆编造当前赛程、赔率、阵容、伤停、天气或盘口变化。
- 只使用公开/授权来源，不注册未知免费 API、不猜 key、不登录抓取或绕过访问控制。
- 概率倾向不等于可购买玩法；确认不可售的市场不得进入购买方案，解析失败则标为未验证而非不可售。
模型首先这样做
优先使用统一入口：
python3 <skill-dir>/scripts/football_skill.py doctor
python3 <skill-dir>/scripts/football_skill.py route '<用户原始请求>'
随后按 next_action 调用 fetch、report、audit、review 或 backtest。完整接口见 references/agent-interface.md。统一入口只编排现有确定性工具，旧脚本继续兼容。
请求状态机
- 分类为单场、组合、赛后复盘或历史回测。
- 中国竞彩默认
china-lottery；明确海外博彩公司才用international-odds；无可验证赔率时使用analysis-only。 - 当前/未来竞彩先检查本地快照，再自动
fetch。无法确认具体比赛时先做 Fixture Discovery；仍失败才索取最少比赛信息。 - 收集比赛身份、时间、地点、赔率/盘口、球队状态和赛事背景，并记录来源与观察时间。冲突不静默平均，按来源质量处理并降低 Data Confidence。
- 按顺序完成 Probability Analysis → Value Judgment → Reference Purchase Plan。没有真实赔率时停止在概率分析；缺少可售市场时不生成该市场购买项。
- 生成正式主单前运行
audit；阻断项必须保护、降级或移除。 - 完成赛前分析后生成单个自包含 HTML 和预测快照；聊天仅返回 2～4 行摘要及路径。
按需加载
- 工作流和模式：
references/workflow.md - 数据来源、时效与冲突：
references/data-sources.md - 数学模型与参数：
references/math-model.md、references/model-parameters.md - 报告格式：
references/report-templates.md - 降级和停止规则：
references/downgrade-rules.md - 回测校准：
references/backtesting.md
只读当前任务需要的文件。术语不熟悉时再读 references/glossary.md。
分析与购买一致性
- 可分析比赛都给出 xG 先验、有限上下文调整、最终 xG、Poisson 比分集中度、赛果/大小球概率和置信度；缺输入时标明近似并降级。
- 有赔率才计算隐含/去水概率、edge 和 Reference Grade。区分 Model Confidence 与 Data Confidence。
- 小组末轮在动机调整前核实排名、积分、净胜球、出线压力、轮换和潜在路线；缺结构化背景的比赛不得进入“模型最稳”。
- 让球结果必须按
主队进球 + 主队让球数 - 客队进球显式核对。候选比分跨越多个让球结果时，要保护、换市场或降级。 - 比分故事与大小球、让球结论必须一致；若不一致，重新校准或清楚说明只有赔率价值才支持逆向选择。
- 组合按真实概率、数据置信、可售玩法和相关性构造，不强迫固定腿数或 2/16/32/48 元档。
- 比分票最多四场；赛果/让球和大小球方向票最多八场。每项写完整选择与
A × B × C = N 注。 - 如果四串一含明显风险腿，另给移除它的更稳三串一；不足三条低风险腿时给二串一并解释。
What ships with it
46 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- README.md 1.1 KB
- references/agent-interface.md 1.1 KB
- references/backtesting.md 3.8 KB
- references/data-sources.md 15 KB
- references/downgrade-rules.md 6.4 KB
- references/glossary.md 4.6 KB
- references/math-model.md 9.9 KB
- references/model-parameters.md 7.6 KB
- references/report-templates.md 15 KB
- references/workflow.md 22 KB
- schemas/agent-result.schema.json 1.0 KB
- schemas/backtest-sample.schema.json 5.3 KB
- schemas/fixture.schema.json 919 B
- schemas/football-snapshot.schema.json 6.3 KB
- schemas/html-report.schema.json 6.7 KB
- schemas/match-analysis.schema.json 4.2 KB
- schemas/odds.schema.json 1.4 KB
- schemas/portfolio.schema.json 2.8 KB
- schemas/team-context.schema.json 1.3 KB
- scripts/auto_post_match_review.py 29 KB runs code
- scripts/backtest_predictions.py 13 KB runs code
- scripts/build_snapshot_report.py 25 KB runs code
- scripts/competition_context_calculator.py 7.9 KB runs code
- scripts/convert_legacy_prediction_snapshot.py 1.4 KB runs code
- scripts/fetch_match_data.py 22 KB runs code
- scripts/football_skill.py 11 KB runs code
- scripts/grade_calculator.py 7.6 KB runs code
- scripts/handicap_rules.py 3.8 KB runs code
- scripts/implied_probability.py 2.1 KB runs code
- scripts/inspect_snapshot_markets.py 2.6 KB runs code
- scripts/late_update_rules.py 3.3 KB runs code
- scripts/market_grade_calculator.py 5.9 KB runs code
- scripts/match_model_calculator.py 6.2 KB runs code
- scripts/poisson_calculator.py 6.1 KB runs code
- scripts/portfolio_builder.py 11 KB runs code
- scripts/post_match_review.py 11 KB runs code
- scripts/prediction_snapshot_tools.py 11 KB runs code
- scripts/preflight_risk_audit.py 13 KB runs code
- scripts/recent_form_to_xg.py 3.4 KB runs code
- scripts/render_html_report.py 21 KB runs code
- scripts/score_coverage_analyzer.py 6.6 KB runs code
- scripts/select_snapshot_matches.py 3.2 KB runs code
- scripts/team_context_rules.py 3.6 KB runs code
- scripts/validate_inputs.py 30 KB runs code
- scripts/xg_prior_calculator.py 7.7 KB runs code
- scripts/zero_operation_smoke.py 3.1 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 74 lines · 132 tokens per session scan A bc5cf3ba5c7d
football-betting-assistant is a skill published in the GitHub repository cocoCzl/coco-skills (5 stars, last pushed 4d ago), licensed MIT. It adds 132 tokens to every session and 1,449 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
sector-rotation
An analysis framework for comparing industries in the Chinese A-share stock market, using business conditions, price momentum, valuation, and money flows. It produces rankings and higher- or lower-allocation suggestions.
strategy-pivot-designer
Detect backtest iteration stagnation and generate structurally different strategy pivot proposals when parameter tuning reaches a local optimum.
twitter-reader
Read Twitter/X for financial research using opencli (read-only). Use this skill whenever the user wants to read their Twitter feed, search for financial tweets, view bookmarks, look up user profiles, or gather market sentiment from Twitter/X. Triggers include: "check my feed", "search Twitter for", "show my…
chenhao-limit-up
A framework for judging Chinese A-share stocks that have reached the daily price-rise limit, using market mood, sector leadership, and trading momentum.
furusato
A Japanese hometown-tax donation manager for furusato nozei, a system where donations to municipalities can qualify for an income-tax or local-tax deduction. It reads donation receipts, stores donation records, and calculates deduction limits.
reading-receipt
An image-reading workflow for extracting structured information from receipts, invoices, and hometown-tax donation certificates. It can first extract text from PDFs and otherwise read their images.