跳到主内容
智客 ZICQ

技能库 智客分类:Agent 工作流 cx-coding-agents

Cx Coding Agents

当用户询问AI Center编码代理数据,想复制或扩展编码代理仪表板,或询问使用,成本,符号,会话,工具,代码影响,用户,模型,跨度,或Claude Code, Codex, Cursor,双子座CLI,或Copilot CLI的日志时使用此技能.

499 安装量

官方网址:skills.sh

技能介绍

先看中文介绍;官方 description 原文单独保留,不改写 SKILL.md。

做什么

当用户询问AI Center编码代理数据,想复制或扩展编码代理仪表板,或询问使用,成本,符号,会话,工具,代码影响,用户,模型,跨度,或Claude Code, Codex, Cursor,双子座CLI,或Copilot CLI的日志时使用此技能.

何时用

官方 description 未单独写出 Use when。按规范,代理会在用户任务与这段 description 的关键词匹配时激活本技能。

代理如何加载

按 Agent Skills 渐进披露:启动时只加载 name 与 description(约 100 token);任务匹配后才读入整份 SKILL.md 正文;scripts/、references/、assets/ 仅在需要时再读。 本文件正文结构:Coding Agents Skill、Supported Agents、Loading References、Safety、CLI Commands、First Response。

文件分析

文件分析:除 SKILL.md 外,正文引用了 references/claude-code.md、references/promql-guidelines.md、references/metrics-querying.md、references/dataprime-reference.md、references/logs-querying.md、references/codex.md,属于带资源的技能包,这些文件按需再读。

官方 description(原文)

Use this skill when the user asks about AI Center Coding Agents data, wants to reproduce or extend the Coding Agents dashboards, or asks questions about usage, cost, tokens, sessions, tools, code impact, users, models, spans, or logs for Claude Code, Codex, Cursor, Gemini CLI, or Copilot CLI.

Coding Agents SkillSupported AgentsLoading ReferencesSafetyCLI CommandsFirst ResponseTime and FiltersPromQL (Claude Code, Gemini CLI, GitHub Copilot Collector)DataPrime (Codex, Cursor, Copilot CLI direct, Claude Code sessions)Cross-Agent QuestionsAnswer Style

来源分类:skills.sh agent-skill

SKILL.md 与 Agent 调用

官方规范 ↗
name
cx-coding-agents
description
Use this skill when the user asks about AI Center Coding Agents data, wants to reproduce or extend the Coding Agents dashboards, or asks questions about usage, cost, tokens, sessions, tools, code impact, users, models, spans, or logs for Claude Code, Codex, Cursor, Gemini CLI, or Copilot CLI.
  1. 发现技能客户端向 Agent 提供名称与描述目录。
  2. 匹配与调用用户指定或任务匹配后,载入 SKILL.md 指令。
  3. 按需加载按步骤读取参考文档、使用脚本与素材。
指令中引用的文件 · 8
  • references/claude-code.md
  • references/promql-guidelines.md
  • references/metrics-querying.md
  • references/dataprime-reference.md
  • references/logs-querying.md
  • references/codex.md
  • references/spans-querying.md
  • references/cursor.md

以下路径提取自原文;文件是否齐全请以来源仓库中的完整目录为准。

具体调用语法与可用工具以目标 Agent 客户端为准。 查看调用机制说明 ↗

安装这个技能

Skills CLI ↗

先选择目标 Agent 和安装范围,保留技能包的附属文件,安装后检查客户端能否发现该技能。

该技能引用了附属文件,请从来源获取完整目录;仅复制 SKILL.md 可能缺少依赖。

交给 Agent 安装

复制安装指令给支持 Agent Skills 的代理,确认其中的目标目录与客户端匹配。

把 Agent Skill「cx-coding-agents」安装到我的项目:SKILL.md 原文与官方 description 见 https://zicq.com/zh/skills/skl-db4cbad4d4aabc3c-Cx-%E7%BC%96%E7%A0%81%E4%BB%A3%E7%90%86.html
请存为 .cursor/skills/cx-coding-agents/SKILL.md 或 .claude/skills/cx-coding-agents/SKILL.md,frontmatter 的 name 与 description 保持原样,不要改写。
该技能还带 scripts/、references/、assets/ 等文件,请从 https://github.com/coralogix/cx-cli 取完整目录,不要只建一个 SKILL.md。

GitHub 完整包 ↗

终端安装 · Skills CLI

需要 Node.js 与 npx。先查看仓库技能列表,确认实际名称。

npx skills add 'https://github.com/coralogix/cx-cli' --list

npx skills add 'https://github.com/coralogix/cx-cli' --skill 'cx-coding-agents'

CLI 会交互选择目标 Agent,默认安装到项目;用户级安装使用 -g。先通过查看命令核对仓库内容,再用 npx skills list 检查已安装技能。

阅读排版
--- name: cx-coding-agents description: | Use this skill when the user asks about AI Center Coding Agents data, wants to reproduce or extend the Coding Agents dashboards, or asks questions about usage, cost, tokens, sessions, tools, code impact, users, models, spans, or logs for Claude Code, Codex, Cursor, Gemini CLI, or Copilot CLI. metadata: version: "1.0.0" internal: true --- # Coding Agents Skill Use this skill as the entry point for any investigation or data question about AI Center Coding Agents. It identifies which data source each agent uses and tells you which reference files to load before querying. ## Supported Agents | Agent | Data Type | Source Identifier | |---|---|---| | Claude Code | Metrics | `claude_code_*` metric family | | Claude Code (session text) | Logs | `source ai_sessions_claude` | | Codex | Logs | `service.name == 'codex_cli_rs'` or `'codex-app-server'` | | Codex (latency) | Spans | `$l.serviceName == 'codex_cli_rs'` or `'codex-app-server'` | | Cursor | Spans | `$l.serviceName == 'cursor-agent'` | | Gemini CLI | Metrics | `gemini_cli_*` metric family | | Copilot CLI (direct OTel) | Spans | `$l.serviceName == 'github-copilot'` or `tags['otel.scope.name'] == 'github.copilot'` | | GitHub Copilot Collector | Metrics | `github_copilot_*` metric family | --- ## Loading References Load the agent reference first, then the shared query-language files it depends on: | Agent | Agent Reference | Shared References | |---|---|---| | Claude Code (metrics) | `references/claude-code.md` | `references/promql-guidelines.md` + `references/metrics-querying.md` | | Claude Code (sessions) | `references/claude-code.md` | `references/dataprime-reference.md` + `references/logs-querying.md` | | Codex | `references/codex.md` | `references/dataprime-reference.md` + `references/logs-querying.md` + `references/spans-querying.md` | | Cursor | `references/cursor.md` | `references/dataprime-reference.md` + `references/spans-querying.md` | | Gemini CLI | `references/gemini-cli.md` | `references/promql-guidelines.md` + `references/metrics-querying.md` | | Copilot CLI (direct OTel) | `references/copilot-cli.md` | `references/dataprime-reference.md` + `references/spans-querying.md` | | GitHub Copilot Collector | `references/copilot-cli.md` | `references/promql-guidelines.md` + `references/metrics-querying.md` | | Cross-agent comparison | All agent references | All shared references | --- ## Safety All query commands (`cx metrics`, `cx logs`, `cx spans`, `cx dataprime`) are read-only. They never modify data and can be run freely without `--yes`. --- ## CLI Commands | Command | Purpose | When to Use | |---|---|---| | `cx metrics search --name ''` | Find metrics by name | Discover available `claude_code_*`, `gemini_cli_*`, `github_copilot_*` metrics | | `cx metrics query ''` | Instant PromQL query | Point-in-time metric snapshot | | `cx metrics query-range ''` | Range PromQL query (time series) | Claude Code, Gemini CLI, GitHub Copilot Collector trends | | `cx logs ''` | Log query | Claude Code session text, Codex logs | | `cx spans ''` | Span query | Cursor, Codex latency, Copilot CLI direct OTel | | `cx dataprime query ''` | Raw DataPrime query | When you need an explicit `source logs` or `source spans` in the query | **Output format:** append `-o toon` to any command when passing results to an agent or further processing. **Multi-profile:** use `-p ` (repeatable) to fan out across profiles simultaneously. --- ## First Response Identify the requested agent and analysis goal. If any required scope is missing, ask before querying: - **Agent:** Claude Code, Codex, Cursor, Gemini CLI, or Copilot CLI. - **Time range.** - **Optional filters:** application, subsystem, user, model, session, repository. - **Output shape:** quick answer, table, trend, top-N ranking, investigation notes, or query only. > If the user says "GitHub Copilot Codex", clarify whether they mean Codex or Copilot CLI. AI Center has separate Codex and Copilot CLI dashboards. Use the same data source as the corresponding dashboard whenever possible: - **Claude Code metrics:** PromQL over `claude_code_*` metrics. - **Claude Code session text:** DataPrime over `ai_sessions_claude` logs. - **Codex:** DataPrime over Codex logs for tokens, sessions, models, users, and tools; spans only for `run_turn` latency. - **Cursor:** DataPrime over Cursor spans. Cursor reports prompt length and file edits, not provider token counts. - **Gemini CLI:** PromQL over `gemini_cli_*` metrics. - **Copilot CLI direct OTel:** DataPrime over Copilot spans. - **GitHub Copilot Collector:** PromQL over `github_copilot_*` org/user/billing metrics when that collector data is present. --- ## Time and Filters ### PromQL (Claude Code, Gemini CLI, GitHub Copilot Collector) Use `increase(metric[])` for counters over the selected window. Common label filter pattern: ```promql {user_email="",model="",cx_application_name="",cx_subsystem_name=""} ``` - Use `sum by (...)`, `count by (...)`, or `topk(N, ...)` for grouping. - For chart trends, use an interval window based on the selected range rather than the full range. ### DataPrime (Codex, Cursor, Copilot CLI direct, Claude Code sessions) - Add `| filter ...` clauses for user, model, application, subsystem, session, and operation. - Prefer `groupby ... aggregate ...` for dashboard-like tables. - For spans: app/subsystem labels are `$l.applicationName` and `$l.subsystemName` (mixed case). - For logs: app/subsystem labels are `$l.applicationname` and `$l.subsystemname` (lowercase). App/subsystem filter pattern: ```text | filter $l.applicationName == '' && $l.subsystemName == '' ``` --- ## Cross-Agent Questions When comparing agents: - **Normalize units** before comparing: cost in USD, tokens as input/output/total, runtime in milliseconds, sessions as distinct conversations. - **State which signals are not equivalent.** Cursor prompt length is not a provider token count. Direct Copilot CLI users are pseudonymous IDs (`enduser.pseudo.id`), not email addresses. GitHub Copilot Collector users may be logins, names, or emails. - **Prefer per-agent sections** over a single merged table when dimensions differ. - For "which agent is most used": compare sessions and users first, then tokens or span counts with caveats. - For "which agent costs the most": Claude Code has first-class USD metrics. Copilot direct spans expose `github.copilot.cost`. Collector billing metrics expose net/gross/discount amounts by organization/SKU. Do not merge Copilot span cost and collector billing without explaining the source and unit difference. --- ## Answer Style Always include: - The agent and data source used. - The query or a compact query excerpt. - The time range and filters applied. - A plain-language interpretation of the result. - Caveats about empty data, approximate counts, pseudonymous users, missing labels, or non-comparable metrics. Do not invent fields not listed in the agent reference files. If the user asks for data an agent does not emit, say what is available and propose the nearest supported query.

相关技能

Agent 工作流

Skill Creator

创造有效技能指南。 当用户想创造出新的技能(或更新现有的技能),以专业知识,工作流程,或工具集成来扩展克洛德的能力时,应该使用这种技能.

Agent 工作流

Clawdhub

使用ClawdHub CLI搜索,安装,更新并发布从taladhub.com的代理技能. 需要获取苍蝇上的新技能时使用,将安装的技能同步到最新版本或特定版本,或者发布 npm-instainddhub CLI 的新/更新的技能文件夹.

Agent 工作流

Agent Team Orchestration

管弦乐团多代理团队,任务设定周期,交接协议,审查工作流程. 使用时间: (1)建立2+特派员队伍,具有不同专业,(2)确定任务路线和生命周期(收录框_ spec_建设_审查_完成),(3)在特派员之间制定交接协议,(4)建立审查和质量关口,(5)管理特派员之间的交流和文物共享.

Agent 工作流

Superpowers

Spec-first,TDD,子代理驱动的软件开发工作流程. 当:(1)构建任何新功能或应用——触发脑暴_计划_子代理执行回路,(2)调试出一个bug或测试失败——触发系统性的根起过程,(3)用户说"让我们构建","帮助我计划","我想添加X",或"这个被打破",(4)完成一个功…