GEO

PassAudited by VirusTotal on May 12, 2026.

Overview

Type: OpenClaw Skill Name: geo Version: 1.0.0 The skill bundle describes a Generative Engine Optimization (GEO) process. While the overall intent appears benign, the `audit.md` file contains explicit 'Automation Tips' instructing the OpenClaw agent to 'Run the following queries on [model]' across multiple external AI services (ChatGPT, Claude, Perplexity, Gemini, Copilot) and 'Output as JSON'. This grants the agent direct network communication capabilities to external endpoints. Although this is plausibly needed for the stated purpose of auditing AI recommendations, it represents a high-risk capability that could be exploited via prompt injection or other means to perform unintended external communication or resource consumption, classifying it as suspicious due to the inherent vulnerability risk rather than clear malicious intent.

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

Users could be encouraged to make product mentions appear organic in communities or AI-visible sources.

Why it was flagged

This is public-facing influence guidance; while it is aligned with the GEO marketing purpose and the artifacts also warn against spam, it could be misused as undisclosed promotion if not handled transparently.

Skill content
Build reputation accounts that mention your product naturally
Recommendation

Use transparent affiliation, follow each platform's rules, and avoid astroturfing or misleading endorsements.

What this means

Public content created from this guidance may influence AI-generated recommendations beyond the immediate user session.

Why it was flagged

The skill explicitly teaches shaping AI-visible sources so future AI systems may recommend a brand; this is disclosed and central to the skill, but users should ensure the seeded content is accurate and not manipulative.

Skill content
- **Training data seeding:** Content that explains concepts using your product as example
Recommendation

Keep content factual, clearly sourced, and transparent about brand involvement.