Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Ultrahuman (OpenClaw)

v0.1.2

Fetch and summarize Ultrahuman Ring/CGM metrics inside OpenClaw using the Ultrahuman MCP server (via mcporter). Use when the user asks about Ultrahuman data such as sleep score, total sleep, sleep stages, HR/HRV/RHR, steps, recovery index, movement index, VO2 max, or wants a daily/weekly Ultrahuman summary.

2· 2k·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
high confidence
Purpose & Capability
The SKILL.md and script clearly implement fetching Ultrahuman data via an Ultrahuman MCP server invoked through mcporter; that matches the skill description. However registry metadata lists no required env vars or binaries even though the script requires ULTRAHUMAN_AUTH_TOKEN, ULTRAHUMAN_USER_EMAIL and mcporter (and the SKILL.md suggests building a Node/Bun-based MCP). The metadata omission is an inconsistency.
Instruction Scope
Runtime instructions are narrowly focused: build/run the Ultrahuman-MCP server, configure mcporter, then call mcporter to fetch metric JSON and summarize it. The script also attempts to read ~/.openclaw/openclaw.json and uses OPENCLAW_WORKSPACE/default workspace paths to find mcporter config — relevant to OpenClaw but worth noting because it reads a user config file for credentials. The MCP server you run will execute network calls to Ultrahuman as part of normal operation.
Install Mechanism
There is no formal install spec (lowest-risk packaging), but the SKILL.md instructs building and running a third‑party repo (Monasterolo21/Ultrahuman-MCP) using bun/node and configuring mcporter. Building and running unvetted third‑party code is a risk the user must evaluate; the skill itself does not install artifacts automatically.
!
Credentials
The skill requires sensitive credentials (ULTRAHUMAN_AUTH_TOKEN and ULTRAHUMAN_USER_EMAIL) but registry metadata does not declare any required env vars or primary credential — a clear mismatch. The script will also read ~/.openclaw/openclaw.json for env.vars fallback, which may expose stored credentials if present. Requested credentials are reasonable for the stated purpose, but the metadata omission and implicit config-file lookup reduce transparency.
Persistence & Privilege
The skill is not marked always:true, is user-invocable, does not request persistent system changes, and does not modify other skills' configs. It runs as an on-demand local script that calls mcporter, which is appropriate for this functionality.
What to consider before installing
This skill appears to do what it says (call an Ultrahuman MCP server via mcporter and summarize the JSON), but there are important caveats to review before installing/using it: - Metadata mismatch: the registry entry declares no required env vars or binaries, but the SKILL.md and script require ULTRAHUMAN_AUTH_TOKEN, ULTRAHUMAN_USER_EMAIL and mcporter (and building the Ultrahuman-MCP repo requires bun/node). Treat the metadata as incomplete and verify requirements manually. - Inspect the MCP code: the SKILL.md tells you to build and run a third‑party repo (https://github.com/Monasterolo21/Ultrahuman-MCP). Building and running that code means executing unvetted Node/Bun code on your machine; review that repository for unexpected network calls, telemetry, or data exfiltration before running it. - Credentials handling: provide only the minimum-scoped token you can. The script will also attempt to read ~/.openclaw/openclaw.json for env.vars — check that file for any other secrets you don’t want accessed and remove or isolate them if necessary. - mcporter and config paths: mcporter executes the MCP server process. Ensure your mcporter config points to the intended local binary (absolute path) and not to an attacker-controlled executable. Prefer explicit --mcporter-config and absolute paths when running the script. - If you need assurance: ask the publisher to update registry metadata to declare required env vars and binaries, and/or provide a signed release of the MCP server. If you don’t trust building external code, do not run the MCP process on sensitive systems. Given these mismatches and the fact the skill executes a separate MCP process you must supply/build, treat the package as suspicious until you verify the MCP repository and your local config/credentials.

Like a lobster shell, security has layers — review code before you run it.

latestvk970q4400cfhpe2xarczax6z2x80dy93

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments