HLE Tunnel

Security checks across malware telemetry and agentic risk

Overview

This is a coherent remote-tunnel skill, but it intentionally exposes selected local or agent services through HLE, so users should be careful about what they share and with whom.

Install this only if you trust HLE as a tunnel provider and need remote access to a local service. Use specific SSO allowlists, avoid unauthenticated tunnels for private services, protect HLE_API_KEY, review active tunnels/share links, and stop any background or systemd tunnel when it is no longer needed.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Risk analysis

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.

#
ASI02: Tool Misuse and Exploitation
Medium
What this means

If the wrong local service or recipient is chosen, someone could gain remote access to private dashboards, notebooks, admin panels, or the agent UI.

Why it was flagged

The skill can expose the agent's local Control UI to a public HLE URL and grant other users access. This is purpose-aligned and disclosed, but it is a high-impact network action.

Skill content
hle expose --service http://localhost:18789 --label my-agent ... --allow friend@gmail.com
Recommendation

Only expose services you intend to share, use SSO with specific allowed emails, avoid '--auth none' for private services, and review active tunnels regularly.

#
ASI03: Identity and Privilege Abuse
Medium
What this means

Anyone with the HLE API key may be able to manage tunnel access for the account.

Why it was flagged

The skill requires an HLE account credential to create and manage tunnels. This is expected for the service, but the credential is sensitive.

Skill content
Run `hle auth login` to save the key ... or set the `HLE_API_KEY` environment variable
Recommendation

Protect the HLE API key, avoid pasting it into logs or shared commands, rotate it if exposed, and use the least-privileged key options HLE provides.

#
ASI04: Agentic Supply Chain Vulnerabilities
Low
What this means

The actual tunnel behavior depends on the upstream HLE CLI or Docker image you install.

Why it was flagged

The skill depends on installing an external HLE client. That is normal for this integration, but the external package code is not included in the provided artifacts.

Skill content
brew | formula: hle-world/tap/hle-client ... uv | package: hle-client
Recommendation

Install from the documented HLE sources, keep the client updated, and verify package/image provenance if using this in a sensitive environment.

#
ASI07: Insecure Inter-Agent Communication
Medium
What this means

Data shown by the exposed local service may become accessible through HLE to the configured users or share-link holders.

Why it was flagged

Local service traffic and UI access are routed through the HLE gateway to allowed remote users. This is the intended function, but it creates a sensitive data boundary.

Skill content
The command runs in the foreground and prints the public URL ... Anyone you `--allow` can log in via Google or GitHub SSO
Recommendation

Expose only the minimum needed service, prefer named SSO users over broad links, revoke temporary links after use, and consider whether the service contains private data.

#
ASI10: Rogue Agents
Low
What this means

A tunnel could remain available longer than intended if started as a background service.

Why it was flagged

The examples show optional background and systemd operation, which can keep a tunnel active after the initial task. This is disclosed and user-directed.

Skill content
nohup hle expose --service http://localhost:18789 --label my-agent > /dev/null 2>&1 & ... Restart=always
Recommendation

Use background or systemd mode only when persistent access is desired, document it, and stop or disable the service when finished.