← 返回 Skills 市场
Analytix402
作者
Analytix402
· GitHub ↗
· v0.1.0
602
总下载
0
收藏
3
当前安装
1
版本数
在 OpenClaw 中安装
/install analytix402
功能描述
Monitor and control your AI agent’s API and LLM usage with real-time spend tracking, budget limits, duplicate detection, and alerts.
使用说明 (SKILL.md)
Analytix402\r
\r Monitor, control, and optimize your AI agent's API spend and LLM costs in real-time.\r \r
Description\r
\r Analytix402 gives your OpenClaw agent financial visibility and guardrails. Track every API call, LLM invocation, and x402 payment your agent makes. Set budget limits, detect duplicate purchases, and get alerts before costs spiral.\r \r What it does:\r
- Tracks all outbound API calls and x402 payments automatically\r
- Monitors LLM token usage and costs across OpenAI, Anthropic, and other providers\r
- Enforces daily budget limits and per-call spend caps\r
- Detects duplicate API purchases to prevent waste\r
- Sends heartbeats so you know your agent is alive and healthy\r
- Provides a real-time dashboard at analytix402.com\r \r
Configuration\r
\r
# Required\r
ANALYTIX402_API_KEY: ax_live_your_key_here\r
\r
# Optional\r
ANALYTIX402_AGENT_ID: my-openclaw-agent\r
ANALYTIX402_BASE_URL: https://analytix402.com\r
ANALYTIX402_DAILY_BUDGET: 50.00\r
ANALYTIX402_PER_CALL_LIMIT: 5.00\r
ANALYTIX402_TRACK_LLM: true\r
```\r
\r
## Tools\r
\r
### analytix402_spend_report\r
Get a summary of your agent's spend — total cost, breakdown by API and LLM provider, and efficiency score.\r
\r
### analytix402_set_budget\r
Set or update the daily budget limit for this agent session.\r
\r
### analytix402_check_budget\r
Check remaining budget before making an expensive API call.\r
\r
### analytix402_flag_purchase\r
Flag a potential duplicate or unnecessary purchase for review.\r
\r
## Tags\r
\r
monitoring, analytics, budget, x402, payments, observability, cost-tracking\r
安全使用建议
Do not supply any API keys or enable this skill for production agents until the author clarifies how telemetry is captured and transmitted. Ask the publisher to: (1) provide an implementation or manifest that shows the exact endpoints, schemas, and minimal data fields sent; (2) declare required environment variables in the registry metadata; (3) publish a privacy/security policy and the analytix402.com ownership information; and (4) explain how it hooks into the agent (does it read request payloads, intercept network calls, require provider keys, or expect the agent to forward events?). If you must test, do so with a throwaway agent running in an isolated environment, monitor outbound network traffic to the listed domain, and avoid giving any provider credentials or sensitive payloads. The current package is internally inconsistent and should be treated with caution.
功能分析
Type: OpenClaw Skill
Name: analytix402
Version: 0.1.0
The skill bundle describes a cost-tracking and monitoring service for AI agents, Analytix402. The `SKILL.md` transparently outlines its functionality, which includes tracking API calls, LLM usage, and sending this data, along with heartbeats, to `analytix402.com`. While this involves outbound network communication and data transfer, it is the explicit and stated purpose of the skill, requiring an API key for configuration. There is no evidence of malicious intent such as unauthorized data exfiltration (e.g., unrelated secrets), malicious execution, persistence mechanisms, or prompt injection attempts to subvert the agent's intended behavior.
能力评估
Purpose & Capability
The description promises automatic tracking of every outbound API call, LLM token usage across providers, duplicate-purchase detection, and heartbeats. But the published package is instruction-only with no code, no install hooks, and the registry metadata lists no required environment variables or permissions. A monitoring/control capability of this form would legitimately need either agent-level hooks, additional declared permissions, or clear instructions for how the agent should forward telemetry; none are present.
Instruction Scope
SKILL.md instructs the agent to use ANALYTIX402_API_KEY and mentions a base URL and tools (analytix402_spend_report, analytix402_set_budget, etc.), but provides no concrete API endpoints, request schemas, or guidance for what data is collected/sent. It also claims to 'track all outbound API calls' and 'monitor LLM token usage' without explaining how to capture those events. This vagueness grants broad discretion to the agent and could result in unexpected data exfiltration if implemented later.
Install Mechanism
No install spec and no code files — lowest install-surface risk from static analysis perspective. However, because there is no implementation included, there is also no way to verify what would actually run. The lack of an install makes the skill easier to audit but also means the SKILL.md alone determines behavior; static scanner had nothing to analyze.
Credentials
SKILL.md lists ANALYTIX402_API_KEY as a required value, but the registry metadata declares no required env vars or primary credential — an explicit mismatch. Additionally, the claimed capability to monitor other providers (OpenAI, Anthropic) would typically require access to those providers' credentials or to the agent's request stream; neither is requested or explained. Requiring an external API key to receive telemetry about potentially sensitive requests is proportionate only if the manifest and privacy/data-handling policies are explicit — they are not.
Persistence & Privilege
always is false and the skill is not force-enabled. The skill can be invoked autonomously (default), which is normal, but combined with the described telemetry/forwarding behavior this increases the blast radius (the agent could autonomously send outgoing request data to analytix402.com). This is not a hard misconfiguration by itself, but it heightens the need to verify what data is sent and to whom.
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 analytix402 - 安装完成后，直接呼叫该 Skill 的名称或使用
/analytix402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
v0.1.0
Initial release of Analytix402, providing real-time API spend and LLM cost monitoring for OpenClaw agents.
- Tracks all API calls, LLM usage, and x402 payments automatically.
- Supports daily budget limits, per-call spend caps, and duplicate purchase detection.
- Sends agent health heartbeats and offers alerts before overspending.
- Introduces tools for spend reports, budget management, and purchase flagging.
- Provides a live dashboard at analytix402.com for real-time visibility.
元数据
常见问题
Analytix402 是什么？
Monitor and control your AI agent’s API and LLM usage with real-time spend tracking, budget limits, duplicate detection, and alerts. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 602 次。
如何安装 Analytix402？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install analytix402」即可一键安装，无需额外配置。
Analytix402 是免费的吗？
是的，Analytix402 完全免费（开源免费），可自由下载、安装和使用。
Analytix402 支持哪些平台？
Analytix402 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 Analytix402？
由 Analytix402（@analytix402）开发并维护，当前版本 v0.1.0。
推荐 Skills