Zero Token

PendingVirusTotal audit pending.

Overview

No VirusTotal analysis has been recorded yet. File reputation checks will appear here once the artifact hash has been scanned.

Findings (0)

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

Provider API keys added to the gateway could be abused if the gateway or local machine is compromised.

Why it was flagged

The skill expects the user to enter API keys for external LLM providers. This is necessary for the stated gateway purpose and the script warns against production keys, but the credentials are still sensitive.

Skill content
Add free API keys ... ⚠️  Use free-tier keys. Do NOT use production API keys.
Recommendation

Use disposable or free-tier keys only, avoid production or billing-enabled keys, and rotate keys if you uninstall or stop trusting the gateway.

What this means

Prompts, files, or user data sent to the fallback model may be processed by whichever free provider is selected.

Why it was flagged

The skill routes model traffic through a gateway to multiple third-party LLM providers, and the documentation correctly notes provider-specific privacy and retention policies.

Skill content
Free-Way gateway aggregates 13+ free LLM providers behind a single `/v1` endpoint... Every provider has its own privacy policy and data retention. Review each provider's terms before routing sensitive data.
Recommendation

Do not route sensitive data until you have reviewed and accepted the privacy terms for each configured provider.

What this means

Installing the skill can execute third-party gateway code and npm packages on the local machine after the user runs the setup script.

Why it was flagged

The setup script downloads a third-party repository and installs npm dependencies. The Free-Way commit is pinned, which reduces risk, but the downloaded code and dependencies are not part of the supplied artifacts.

Skill content
git clone --depth 1 "$FREE_WAY_REPO" "$FREE_WAY_DIR" ... git checkout FETCH_HEAD ... npm install --omit=dev 2>/dev/null || npm install
Recommendation

Review the pinned Free-Way commit and its dependency lockfile before running setup, and avoid changing the pinned commit without review.

NoteHigh Confidence
ASI10: Rogue Agents
What this means

The local gateway may keep running after setup until the user stops it.

Why it was flagged

The script starts the gateway as a background process. This is disclosed and includes stop/uninstall commands, so it is not hidden persistence.

Skill content
nohup npm start > "$FREE_WAY_DIR/free-way.log" 2>&1 & ... To STOP the gateway: pkill -f 'node.*free-way'
Recommendation

Stop the service when it is not needed and remove `$HOME/free-way` if you uninstall the skill.

NoteMedium Confidence
ASI01: Agent Goal Hijack
What this means

Using the skill may change how the agent responds, not just how it routes model calls.

Why it was flagged

The skill instructs the agent to apply a separate SOUL behavior profile, which can affect the agent's identity, tone, memory behavior, and upgrade suggestions.

Skill content
T:apply_soul|from:SOUL.md ... T:apply_genes|all
Recommendation

Review SOUL.md before enabling the skill and remove or modify behavior genes that do not match your desired agent behavior.

What this means

End users may assume they are still using the primary model unless the application clearly discloses fallback provider behavior.

Why it was flagged

The marketing emphasizes seamless or invisible failover while also claiming transparency. This appears to mean service continuity, but users should not confuse seamless fallback with unchanged privacy or provider handling.

Skill content
Your users won't even notice the switch... Zero Token never hides the switch.
Recommendation

Disclose fallback routing in your own product or agent policy, especially when prompts may go to free-tier providers.