Apple Health

ReviewAudited by ClawScan on May 1, 2026.

Overview

This skill is coherent and privacy-conscious, but it will run a third-party Apple Health MCP package on your local health export and save local setup/query notes.

Before installing, make sure you trust the third-party Apple Health MCP package and the MCP client you will connect it to. Point HEALTH_DATA_DIR only at the specific export folder you want analyzed, prefer bounded queries, and periodically review or delete the local ~/apple-health/ memory if it contains sensitive summaries.

Findings (3)

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 npm package will execute locally and is expected to access the Apple Health export directory.

Why it was flagged

The skill directs users to run a third-party npm MCP server, without a pinned package version, as the main integration mechanism.

Skill content
"command": "npx", "args": ["@neiltron/apple-health-mcp"]
Recommendation

Install only if you trust the MCP package; consider reviewing its source/provenance and pinning a known-good version where your client supports it.

What this means

Health summaries, export paths, and analysis notes may remain on disk and be reused in later sessions.

Why it was flagged

The skill stores persistent local memory, including query logs and summaries that may contain health-related context.

Skill content
Memory lives in `~/apple-health/` ... `query-log.md` # Reusable SQL/report prompts and known-good outputs
Recommendation

Keep stored notes minimal, avoid saving raw records, and delete or redact ~/apple-health/ if you no longer want the agent to retain this context.

What this means

Sensitive health results may be visible to whichever MCP-compatible client or agent workspace you connect.

Why it was flagged

Health data is accessed through an MCP client/server flow, so query outputs can enter the connected client or agent conversation context.

Skill content
Then confirm the MCP client ... Run `health_schema` ... Run one bounded query from `query-recipes.md`
Recommendation

Use only trusted MCP clients and avoid connecting this server in shared or untrusted workspaces unless you are comfortable exposing health summaries there.