← 返回 Skills 市场
nftechie

Transition MCP

作者 nftechie · GitHub ↗ · v1.0.0
cross-platform ⚠ suspicious
854
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install transition-mcp
功能描述
AI-powered multisport coaching — get personalized workouts, training plans, and performance analytics for running, cycling, swimming, and triathlon.
安全使用建议
This skill appears to implement what it claims (calls Transition API endpoints and provides MCP server tools), but the package metadata is incomplete: it does not declare the TRANSITION_API_KEY env var or the need for the Go toolchain even though both are required by the included code. Before installing or running: 1) Verify you trust https://transition.fun and the skill author (source is 'unknown' in registry). 2) Expect to supply TRANSITION_API_KEY (starts with tr_live_); store it securely (avoid committing to repo). 3) If you plan to run the MCP server locally, ensure you have the Go toolchain and inspect dependencies (github.com/modelcontextprotocol/go-sdk). 4) If you only want unauthenticated features, use the free /wod endpoint (no API key). 5) Ask the publisher to correct the metadata (declare required env vars and required binaries) or provide a signed release binary to simplify verification. If you are uncomfortable with building/running third‑party code that performs network calls to an external API, do not install and instead use the unauthenticated curl examples against the public endpoints.
功能分析
Type: OpenClaw Skill Name: transition-mcp Version: 1.0.0 The skill bundle is a client for the Transition API, providing AI-powered multisport coaching features. All network communication is directed to the legitimate domain `https://api.transition.fun`. The `SKILL.md` provides clear instructions for the AI agent, including 'Tips for Agents', which are benign and aim to guide responsible agent behavior, not to perform prompt injection for malicious purposes. API keys are handled securely via environment variables, and there is no evidence of data exfiltration, malicious execution, persistence mechanisms, or obfuscation across any of the analyzed files.
能力评估
Purpose & Capability
The SKILL.md and included Go code clearly expect a TRANSITION_API_KEY (and optionally TRANSITION_API_URL) and provide an MCP server implementation; however the registry metadata lists no required environment variables or binaries. The presence of mcp/*.go files implies you need a Go toolchain and the modelcontextprotocol go-sdk to build/run — these are not declared in metadata.
Instruction Scope
Runtime instructions (SKILL.md, README, examples) stay within the advertised purpose: calling https://api.transition.fun endpoints, polling generation status, and using the API key for authenticated calls. There are no instructions to read unrelated system files or exfiltrate data beyond the Transition API. The README does show how to add the API key to a shell profile or an MCP config file (which stores the key locally).
Install Mechanism
There is no formal install spec (instruction-only), but source code is bundled (Go). Building the MCP server requires 'go build' and pulling dependencies (github.com/modelcontextprotocol/go-sdk/mcp). The absence of an install spec/binary declaration is a packaging inconsistency (not necessarily malicious) and means a user who follows the README will compile/run code locally.
Credentials
Only one real credential is required by the code: TRANSITION_API_KEY (and an optional TRANSITION_API_URL). The registry metadata incorrectly lists no required env vars or primary credential. That omission is disproportionate and misleading — the skill will only function for personalized features if you provide the API key, and the README/SKILL.md direct you to store it in shell/profile or an MCP config file.
Persistence & Privilege
The skill does not request 'always: true' and does not modify other skills or system-wide settings. Running the MCP server registers tools/resources for an MCP client, but that is standard for this type of skill and is limited to the local process when you run it.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install transition-mcp
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /transition-mcp 触发
  4. 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
v1.0.0
Initial release of transition-mcp: AI-powered multisport coaching API. - Provides structured, personalized workouts and training plans for running, cycling, swimming, and triathlon. - Offers a free Workout of the Day (WOD) endpoint—no authentication required. - Authenticated features include workout scheduling, plan adaption, performance analytics (CTL, ATL, TSB), and AI coach chat. - Enables integration with Garmin, chat history retrieval, and athlete profile access. - Includes clear rate limits for free and paid tiers. - Usage tips are provided for agents, including how to check fatigue and when to adapt plans.
元数据
Slug transition-mcp
版本 1.0.0
许可证
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Transition MCP 是什么?

AI-powered multisport coaching — get personalized workouts, training plans, and performance analytics for running, cycling, swimming, and triathlon. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 854 次。

如何安装 Transition MCP?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install transition-mcp」即可一键安装,无需额外配置。

Transition MCP 是免费的吗?

是的,Transition MCP 完全免费(开源免费),可自由下载、安装和使用。

Transition MCP 支持哪些平台?

Transition MCP 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 Transition MCP?

由 nftechie(@nftechie)开发并维护,当前版本 v1.0.0。

💬 留言讨论