agent-analytics-report
by community
Use this page as a decision snapshot for agent-analytics-report: trust signal, install momentum, real user feedback, and high-intent related pages you can compare next.
Description
Agent 用量分析报告生成器(支持日/周/月/年)。从本地数据源(traces、workbuddy.db、usage-log.json、会话目录)采集 Agent 使用数据,一键生成可读、可分享的多格式报告。数据源用 --source 切换:workbuddy(默认,读 ~/.workbuddy/)/ claude-code(读 ~/.claude/projects/ 的 JSONL 会话日志)/ codex(读 ~/.codex/sessions/ 的 rollout JSONL)/ qwenwork(读千问办公 ~/.qwenworkcn/ 的转录 + 调用日志 + agents.db,**在千问办公里跑就用它**)/ dumate(按 WorkBuddy 布局读取数据目录的通用入口,支持 DUMATE_HOME 覆盖;**数据归属未证实,报告内容可能完全是 WorkBuddy 的,勿当作百度搭子用量引用**);更多 Agent 可扩展(详见 docs/ADAPTERS.md)。 触发方式:当用户说「生成(Agent)周报 / 月报 / 年报 / 日报」「帮我出一份本周使用报告」「统计下这个月的 token 消耗」等、**且语境指向 Agent 用量 / 成本 / Token 时**触发,无需手动指定参数;若用户要的是通用周报 / 月报 / 财务报表,应先澄清数据源再采集。也可用 --period / --days / --start / --end 自定义周期与日期范围。 报告包含: - Token 消耗与成本:按实际计费模型对账,跟后台账单一致(⚠️ 千问办公等服务端不回传 token 的数据源除外:该源 token 为字符估算、金额一律以 `tokens_only` 呈现,估算值不可用于对账,见数据源表);每日趋势、缓存占比、成本货币化 - 任务与技能:这段时间主要在干哪类活、哪些技能被反复调用 - 自动化运行:每个任务跑了几次、成功还是失败、失败浪费了多少钱 - 异常检测:成本 + Token 双口径,免费期高流量也不漏报;脏数据显式警告 输出格式:Markdown(默认)/ HTML(可交互图表,浅色/深色自适应)/ JSON。 **隐私与权限(分阶段)**:① **报告生成流程默认只读 + 默认离线**:默认零网络、不上传任何数据、不递归扫 `~/.workbuddy`;**报告默认只在终端 / stdout 与内存中渲染,不写文件**。两个独立 opt-in(非默认行为):当你显式传 `--output`/`-o <路径>` 时,才把报告写入你指定的本地文件(路径由你控制);当你显式开启 `--task-classifier llm`(须加 `--task-classifier-llm-ack` 或环境变量确认)时,才会把会话派生文本发往你指定的本地 / 自有 LLM 端点做任务分类。② **自定义模型注入(可选、独立流程)**:仅在你明确发起时(说一句"把我的自定义模型加进报告"),Agent 才会在 `~/.workbuddy` 内做有界 `custom-local:*` 文本匹配,逐个问你单价;只有你确认单价后,才写入本地 `pricing.local.json`(补自定义模型单价)——**此文件不进发布包、升级不会被覆盖**。③ `scripts/task_rules.json`(调任务分类权重)**是发布资产、会被升级覆盖**,改它前请自行备份。逐条联网行为见正文「隐私与联网行为(完整清单)」。 限制说明:数据来自本机,需启用 trace 记录;成本按模型公开单价估算,以实际账单为准;免费/限免期内成本口径参考意义有限,报告会自动标注并引导查看 Token 口径。
Install agent-analytics-report
Run this in your OpenClaw agent to add agent-analytics-report from the ClawHub registry.
openclaw skills install agent-analytics-reportRequires ClawHub registry access. Review the security analysis below before installing.
Security Analysis
Open Source
Code is publicly available for audit.
Community Verified
Reviewed by the ClawHub community.
Community Reviews
Real user ratings only — separate from the editorial assessment and ClawHub signal.
Installed this skill? Sign in and leave the first review.
Save the skill now, come back after testing it, and help the next person choose with a quick review.
Related skills
Frequently asked questions
Is agent-analytics-report safe to install?
agent-analytics-report has a SkillsReview security score of 50/100. It is open source and community-verified on ClawHub. Check the full Security Analysis on this page before installing.
How much does agent-analytics-report cost?
agent-analytics-report is free to install for OpenClaw via ClawHub.
What are the best alternatives to agent-analytics-report?
You can compare agent-analytics-report side by side with similar OpenClaw skills on the SkillsReview comparison page to find the best fit for your workflow.
How do I install agent-analytics-report?
Install agent-analytics-report from ClawHub at clawhub.ai/skills/agent-analytics-report, or use the install action on this page to copy the command for your OpenClaw agent.
Community Signal
Historical movement
Timeline plus trend snapshots for security, reviews, and reputation tilt.
Trend Charts
30 / 90 / 180 day snapshots for ranking movement and security-score movement.
Last updated unknown UTC
Submit your review
Share your experience and help others find the best skills.
Newsletter
Stay updated on agent-analytics-report and the wider SkillsReview ecosystem
Get the weekly Top 5, fresh security alerts, and newly hot skills by email. You can unsubscribe from any newsletter email in one click.