← 返回 Skills 市场
agent-analytics
作者
Danny Shmueli
· GitHub ↗
· v4.0.37
· MIT-0
7977
总下载
2
收藏
0
当前安装
61
版本数
在 OpenClaw 中安装
/install agent-analytics
功能描述
Product analytics with your AI agent: set up consent-based tracking, read funnels, paths, retention, experiments, and context, then recommend the smallest growth action using the official Agent Analytics CLI.
安全使用建议
Install only if you are comfortable using the Agent Analytics npm CLI and granting browser-approved access to your analytics account. Before setup, confirm the project goal and event plan, review any tracker/code changes, keep `.openclaw/agent-analytics/config.json` out of git, and avoid storing PII, secrets, or temporary analysis notes in project context.
能力评估
Credentials
The skill may inspect the user's repo, modify instrumentation, use browser/detached auth, and read analytics data including sessions and replays; those are sensitive but proportionate for product analytics and paired with consent-based setup and verification rules.
Install Mechanism
The skill requires running `npx --yes @agent-analytics/[email protected]`, which downloads and executes an external npm package; this is clearly disclosed, pinned to a version, and central to the skill's stated function.
Instruction Scope
Runtime instructions are scoped to Agent Analytics workflows, prefer fixed CLI commands, discourage raw API/curl substitutes, require user-confirmed goals before durable context storage, and explicitly warn against overtracking, PII, secrets, git logs, and unsupported guesses.
Persistence & Privilege
It stores CLI auth under `.openclaw/agent-analytics` in managed runtimes and can write compact product context through `context set`; both are disclosed, scoped to the analytics service, and include guidance not to commit credentials or store PII/secrets.
Purpose & Capability
The skill purpose is product analytics setup and analysis, and the capabilities it requests fit that purpose: installing a tracker, reading funnels/events/sessions/replays, managing project context, and recommending growth actions.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install agent-analytics - 安装完成后，直接呼叫该 Skill 的名称或使用
/agent-analytics触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v4.0.37
Pin the official CLI to 0.5.37, which removes preview upgrade prompts and requires a fresh signed-in full scan.
v4.0.36
Pin CLI 0.5.36 and keep full commands and private context out of payment handoff URLs.
v4.0.35
Pin CLI 0.5.35 with country reporting in stats and events.
v4.0.34
Use CLI 0.5.34, surface replay availability in session analysis, and open linked replay segments by analytics session ID.
v4.0.33
Pin Agent Analytics CLI to 0.5.33.
v4.0.32
Use canonical Agent-Analytics/skills repository URL.
v4.0.31
Redesign Agent Analytics skill package and keep install path on agent-analytics/skills.
v4.0.30
Clarify closed-loop growth recipe uses current CLI commands; no report command in CLI 0.5.28.
v4.0.29
Improve Agent Analytics funnel answers with analytics-oriented diagnosis, metric caveats, and portfolio-aware tactic cards.
v4.0.28
Pin Agent Analytics skill to @agent-analytics/[email protected].
v4.0.27
Use browser approval by default for Claude Code, Codex, Cursor, and local CLI flows; keep detached login for Paperclip, OpenClaw, and issue-based runtimes; pin CLI 0.5.25.
v4.0.26
Remove website scan workflow from public skill setup guidance; clarify pinned CLI use and user-consented project-owned tracking.
v4.0.24
Pin CLI to 0.5.23 and document identity journey lookup by user id or local-only email hash.
v4.0.23
Teach portfolio-scoped cross-project identity stitching with data-link-domains and surface_roles
v4.0.22
- Updated SKILL.md for a broader, more builder-focused positioning: emphasizes multi-surface product analytics, AI builders, and growth loops.
- Clarified and simplified compatibility notes, with improved guidance on login and setup—no API key or chat secrets needed.
- Broadened tags (added "growth" and "builders") for better discovery.
- Clarified the handling of CLI limits: now explicitly instructs to stay on the CLI and guide upgrades, rather than handoff.
- Minor text edits: streamlined English, expanded context on use cases, updated some details for clarity and completeness.
v4.0.21
- Bumped the Agent Analytics CLI version used for all commands to @agent-analytics/[email protected].
- Added extensive documentation on subagent delegation support, including new usage patterns, delegation rules, and recommended splitting strategies for growth audits.
- Included a new "delegation" tag to reflect expanded capability.
- No changes to compatibility or upgrade requirements for standard use cases.
v4.0.20
Teach self-improving project context: read project context before analysis, save durable activation/event/glossary learning, and skip stale observations.
v4.0.19
Add persistent auth storage guidance for OpenClaw and pin CLI 0.5.16.
v4.0.18
Use @agent-analytics/[email protected] and document detached login as a handoff-only flow by default, with polling only via --wait/--poll.
v4.0.17
Pin Agent Analytics CLI examples and runtime instructions to @agent-analytics/[email protected] and refresh public aliases.
元数据
常见问题
agent-analytics 是什么？
Product analytics with your AI agent: set up consent-based tracking, read funnels, paths, retention, experiments, and context, then recommend the smallest growth action using the official Agent Analytics CLI. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 7977 次。
如何安装 agent-analytics？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install agent-analytics」即可一键安装，无需额外配置。
agent-analytics 是免费的吗？
是的，agent-analytics 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
agent-analytics 支持哪些平台？
agent-analytics 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 agent-analytics？
由 Danny Shmueli（@dannyshmueli）开发并维护，当前版本 v4.0.37。
推荐 Skills