← 返回 Skills 市场
Openrouter Usage
作者
Luis Cajigas
· GitHub ↗
· v0.1.0
421
总下载
2
收藏
3
当前安装
1
版本数
在 OpenClaw 中安装
/install openrouter-usage-info
功能描述
Track OpenRouter API spending — credit balance, per-model cost breakdown, and spending projections from OpenClaw session logs.
安全使用建议
This package appears internally consistent with its purpose, but before installing: 1) verify the package source (the registry lists source/homepage as unknown; prefer installing from a trusted GitHub repo or reviewing the code you downloaded), 2) review install.sh to confirm you are comfortable with a CLI being placed in ~/.local/bin and an optional symlink into ~/.openclaw/workspace/skills, 3) understand the tool will read your OpenClaw session logs (~/.openclaw/agents/*/sessions/*.jsonl) and any OpenClaw auth files (~/.openclaw/agents/*/agent/auth.json) to auto-discover keys, and it will send your OpenRouter API key to openrouter.ai when fetching credits — if you are uneasy, create or use a limited/revocable API key, or run the script in a sandbox and inspect its output first.
功能分析
Type: OpenClaw Skill
Name: openrouter-usage-info
Version: 0.1.0
The OpenClaw skill bundle 'openrouter-usage' is a benign utility designed to track OpenRouter API spending. It legitimately accesses the OpenRouter API (openrouter.ai) to fetch credit balance and reads local OpenClaw session logs (`~/.openclaw/agents/*/sessions/*.jsonl`) to provide per-model cost breakdowns. The `install.sh` script performs standard user-level installation, creating a CLI wrapper and an optional symlink for the skill, without any malicious commands or unauthorized system modifications. The `SKILL.md` and `README.md` provide clear, non-malicious instructions and documentation, with no evidence of prompt injection attempts or hidden agendas.
能力评估
Purpose & Capability
The skill claims to report OpenRouter credits and per-model usage from OpenClaw session logs and the code does exactly that: it queries openrouter.ai for credit info and parses ~/.openclaw/agents/*/sessions/*.jsonl for per-model cost. The only required binary is the CLI wrapper it installs (openrouter-usage), which is consistent with the README/SKILL.md.
Instruction Scope
SKILL.md instructs running the provided CLI and documents that the API key is auto-discovered and that session parsing is local (reads ~/.openclaw/agents/*/sessions/*.jsonl). The Python script also only reads those local files and calls the OpenRouter API; it does not reference other system paths or external endpoints beyond openrouter.ai.
Install Mechanism
There is no remote installer: install.sh writes a wrapper to ~/.local/bin and can optionally symlink the skill into ~/.openclaw/workspace/skills — this is a local install and expected. Note: the registry metadata called this an instruction-only skill, but code files and an installer are present in the package (minor inconsistency). Also README suggests cloning a GitHub repo as the install source, but the registry entry lists source/homepage as unknown; verify the repository origin before running the installer.
Credentials
No required environment variables are declared, which matches registry metadata; however the script will optionally use OPENROUTER_API_KEY (env) or keys found in ~/.openclaw/agents/*/agent/auth.json. This is proportionate for its purpose (it needs an OpenRouter key to query credits) but you should be aware the key will be sent to openrouter.ai in an Authorization header.
Persistence & Privilege
The skill does not request always:true. The installer may create a CLI in ~/.local/bin and optionally link the skill into the user's OpenClaw workspace (with explicit prompt). It does not modify other skills or global system config beyond those user-visible local changes.
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 openrouter-usage-info - 安装完成后，直接呼叫该 Skill 的名称或使用
/openrouter-usage-info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 — track OpenRouter API spending from your OpenClaw agent.
- View credit balance, model-by-model cost breakdown, and spending projections.
- Flexible reporting: full report, credits only, or detailed per-model breakdown.
- Filter by timeframe (today, week, month, or custom dates) and agent.
- Supports JSON output and UTC/local timezone filtering.
- No extra dependencies; works locally using your OpenClaw session logs.
元数据
常见问题
Openrouter Usage 是什么？
Track OpenRouter API spending — credit balance, per-model cost breakdown, and spending projections from OpenClaw session logs. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 421 次。
如何安装 Openrouter Usage？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install openrouter-usage-info」即可一键安装，无需额外配置。
Openrouter Usage 是免费的吗？
是的，Openrouter Usage 完全免费（开源免费），可自由下载、安装和使用。
Openrouter Usage 支持哪些平台？
Openrouter Usage 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 Openrouter Usage？
由 Luis Cajigas（@lcajigasm）开发并维护，当前版本 v0.1.0。
推荐 Skills