NotFair Google Ads for OpenClaw

Security checks across malware telemetry and agentic risk

Overview

This appears to be a coherent NotFair/Google Ads integration, but it can access stored NotFair credentials and invoke high-impact Google Ads tools, so users should review approvals carefully.

Install this only if you trust NotFair with your connected Google Ads data. Treat any campaign, budget, bid, keyword, ad, or delete action as money-affecting: review the exact proposed change before approving it, and remove stored tokens/config entries when you no longer need the plugin.

VirusTotal

62/62 vendors flagged this plugin 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 a write-capable tool is approved, it could change campaigns, bids, budgets, keywords, ads, or other Google Ads account state.

Why it was flagged

The skill exposes a broad pass-through to NotFair Google Ads tools, including write-capable operations, while instructing the agent to obtain explicit approval before writes.

Skill content
- notfair_google_ads_tool - call a specific NotFair MCP tool by exact name and JSON arguments.

- Any Google Ads write must be explicitly approved by the user before execution.
Recommendation

Before approving a write, review the exact tool name, arguments, affected account, budget/spend impact, and whether the change can be paused, drafted, or reversed.

#
ASI03: Identity and Privilege Abuse
Medium
What this means

Anyone with access to the stored token/config may be able to access the connected NotFair/Google Ads integration within that token's permissions.

Why it was flagged

The plugin uses delegated NotFair credentials to access the hosted MCP service and stores sensitive token material in OpenClaw plugin configuration.

Skill content
"accessToken": { "label": "Access Token", "sensitive": true },
"apiKey": { "label": "API Key", "sensitive": true, "help": "Optional Bearer token for headless use. Set with openclaw notfair login --token." }
Recommendation

Use a dedicated NotFair token where possible, protect ~/.openclaw/openclaw.json, and run logout or remove the token if you no longer need the integration.

#
ASI05: Unexpected Code Execution
Low
What this means

User-provided analysis scripts may query live advertising data through NotFair, even if they are intended to be read-only.

Why it was flagged

The runScript tool accepts JavaScript source for server-side Google Ads analysis. The artifacts describe this as read-only and central to the diagnostic purpose.

Skill content
description: "JavaScript source. Top-level await is supported by NotFair's runScript tool."
Recommendation

Run scripts only for diagnostics you understand, avoid including unnecessary sensitive data in script code, and reserve write operations for explicitly approved tools.

#
ASI10: Rogue Agents
Low
What this means

After login/setup, the plugin may remain enabled and available to the agent until the user removes it or edits configuration.

Why it was flagged

The plugin can persistently add itself to OpenClaw plugin/tool allow lists. The permission metadata says this is to keep the plugin enabled across restarts, and the change is scoped to this plugin.

Skill content
if (!allow.includes(PLUGIN_ID)) { allow.push(PLUGIN_ID); raw.plugins.allow = allow; }
...
if (!alsoAllow.includes(PLUGIN_ID)) { alsoAllow.push(PLUGIN_ID); raw.tools.alsoAllow = alsoAllow; }
Recommendation

If you stop using NotFair, run the plugin logout/removal flow and check OpenClaw configuration for the openclaw-notfair allow-list entries.