health-sync

SuspiciousAudited by ClawScan on May 10, 2026.

Overview

The skill is mostly coherent for health analysis, but it may sync and store Eight Sleep data for bed occupants or related users, not just the installing user.

Review before installing. Only use this on a trusted host, because it stores provider credentials and detailed health data. If you connect Eight Sleep, verify whether partner/bed-occupant data will be synced and get consent or disable that collection. Also review the external health-sync npm package before uploading the encrypted credential archive.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

The bot could sync, store, and analyze a partner's or household member's sleep/profile data without that person clearly consenting through this skill.

Why it was flagged

The stated skill purpose focuses on the user's health data, but this reference describes ingesting profiles for other bed occupants or related users from Eight Sleep.

Skill content
### `users` (Bed Occupants / Related Users) ... gathered from: current user id, devices.result.leftUserId, devices.result.rightUserId, devices.result.awaySides.*
Recommendation

Restrict Eight Sleep syncing to the authenticated/current user by default, or clearly disclose related-user collection and require explicit consent or an exclusion option.

What this means

Anyone with access to the bot host or backups could potentially access health records or provider tokens.

Why it was flagged

The skill persistently stores provider credentials and a local health database on the bot host, which is expected for sync but sensitive.

Skill content
Expected local working paths: ... workspace/health-sync/.health-sync.creds ... workspace/health-sync/health.sqlite ... Finish flow writes decrypted secrets to local files on the bot host.
Recommendation

Use this only on a trusted host, protect the credential and SQLite files, define retention/deletion practices, and revoke provider access when no longer needed.

What this means

A normal health question may refresh data from connected providers before the answer is produced.

Why it was flagged

The agent is instructed to execute a local CLI sync before analysis; this is disclosed and purpose-aligned, but it can contact providers and update local data automatically.

Skill content
Before any analysis, always run: `npx health-sync sync`
Recommendation

Confirm which providers are enabled and ask for stale-data analysis if you do not want a fresh sync for a particular request.

What this means

The security of credential handling and syncing depends on the external health-sync package and its publisher.

Why it was flagged

The main runtime behavior comes from an external npm package rather than code included in the reviewed skill artifacts.

Skill content
node | package: health-sync | creates binaries: health-sync
Recommendation

Review or pin the npm package version/source before connecting sensitive health accounts.