Case file No.496 · npm · stdio · streamable-http

Mails.ai — Agent Email

ai.mails/agent-email · v0.5.0

A real inbox for AI agents: send, receive and thread email, behind a prompt-injection firewall.

Package ↗Website ↗

Security score

73/100

Good

Normalized over 80 points of available evidence.

Unknown: Maintenance

Score breakdown

Known advisories40/40

0 affecting current version, 0 historical

Source transparency0/15

no public GitHub repository linked

Metadata stability15/15

0 risky metadata change(s) without a version bump

Adoption3/10

414 weekly downloads

Maintenanceunknown

Excluded from the score — no data available.

Known advisories

source: OSV.dev

No advisories found in OSV for this package name. (Not proof of safety.)

Metadata change timeline

rug-pull detection

First snapshot (v0.5.0)Oct 17, 2026

No changes observed yet across 1 snapshot(s).

Tool lists: not published by the registry — unknown.

FAQ

auto-generated from the data above

Is Mails.ai — Agent Email safe?

MCPWatch gives Mails.ai — Agent Email a heuristic security score of 73/100 (good), based on 80 points of available evidence; unknown: Maintenance. This is an automated heuristic, not an audit.

Does Mails.ai — Agent Email have known CVEs or vulnerabilities?

No advisories were found in OSV.dev for the npm package @mailsai/mcp-server. That is not proof of safety.

When was Mails.ai — Agent Email last updated?

The latest registry version is 0.5.0, published Oct 16, 2026. The last repository push is unknown. MCPWatch last checked it on Oct 20, 2026.

What license does Mails.ai — Agent Email use?

Unknown — no public GitHub repository is linked in the registry.

Related servers

same publisher / similar score