360 Secure SkillHub
360 安全技能中心

cricket-player-team-comparison

通过 Crawlora 的 Cricinfo Statsguru/records 接口,先核实赛制、实体 ID 与表结构,再按统一分母和单位对比球员或球队的击球、投球、守备等记录,输出筛选台账、对比表、截断与未知字段说明及按角色的解读,并保留投球回合记号的换算规则。

crawlora-org
crawlora-org
热度 7.65
最近更新 2026-10-05 10:13:56
分类:数据依赖:第三方 API / MCP 服务 / 代码执行应用场景:研究 / 数据处理 / 报告撰写 / API 集成

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add cricket-player-team-comparison
Metadata
name
cricket-player-team-comparison
description
Compare cricket player or team records through Crawlora Cricinfo Statsguru, records, matches, and discovery. Use for format-specific batting/bowling/fielding comparisons with class, innings, dates, opposition, overs notation, and source table definitions preserved.

Cricket player and team comparison

Access and scope

Set CRAWLORA_API_KEY to your Crawlora key. Prefer the listed Crawlora MCP tools when connected; otherwise use the bundled scripts/crawlora.sh, which sends x-api-key to https://api.crawlora.net/api/v1. Read reference/endpoints.md for exact methods and parameters. Check application code as well as HTTP status; payloads are inside data. Stop on authentication errors, back off on 429, and retry a transient upstream failure once. Bound calls/pages by the requested scope and credit budget. Retain source IDs, URLs, source dates, and observation/crawl times separately.

Compare a defined player/team cohort under an explicit match format, season or date window, statistic family, and role. Separate international/domestic and men's/women's or age-group contexts rather than treating every record as comparable.

Discover class, entity, and table scope

  1. Use the records index and team/match/series sources to obtain supported match classes and verified numeric team/player/opposition/ground IDs. A class code selects a format/competition scope; do not infer it from a similar league name. There is no general free-text player-search parameter on Statsguru here.
  2. Statsguru requires class and type. Use only the documented families (batting, bowling, fielding, allround, fow, team, official, aggregate) and view values. Retain each exact filter, season/date bounds, opposition/host/ground context, source URL, and observation time.
  3. Source tables have variable headers/row fields. Interpret displayed labels for that table, not a fixed universal schema. Missing/dash values remain unknown. A player selector changes the table scope; do not assign global rows to that player simply because the request included an identifier.
  4. A Statsguru result is bounded by limit (up to 100), without an exposed arbitrary page argument. Record table coverage. Record paths come from public record discovery; IDs/paths are not interchangeable with canonical match URLs. If rankings are used, preserve their source URL, date, format, and ranking scope.
scripts/crawlora.sh /cricinfo/records/index
scripts/crawlora.sh /cricinfo/teams
scripts/crawlora.sh /cricinfo/stats class=1 type=batting view=year limit=10
# Use returned/verified entity IDs before player/team-specific comparisons.

Normalize rates without changing cricket notation

Compare runs, wickets, matches, innings, balls, not-outs, and opportunities on matching bases. Batting average uses dismissals rather than appearances; bowling average and economy use different numerators/denominators. Calculate rates only when compatible raw counts and table definitions are available, retaining the provider's own rate and rounding separately.

Overs notation is not decimal time: its final component represents balls in an incomplete over. Verify the applicable balls-per-over rule before converting overs into balls or computing economy; do not treat displayed 5.3 as 5.3 decimal overs. Team innings, player batting innings, and bowling spells are different exposures. Do not sum unlike format records, average percentages without event weights, or compare complete-career totals with a recent selected sample.

Return an entity/class/filter ledger, comparison table with denominators and units, selected match/context evidence, truncation and unknown fields, and a role-specific interpretation. Rankings/records are descriptive source observations, not forecasts, injury assessments, or betting recommendations. Strong aggregate figures do not establish performance against unobserved opposition or conditions.