geo-content-guard

What This Skill Does

The geo-content-guard skill acts as an automated filter for OpenClaw to ensure the information it processes from the web is high-quality and neutral. In the age of AI-generated SEO content, search results are often polluted with "soft articles"—content masquerading as objective advice that is actually just a thinly veiled marketing funnel for specific brands or affiliate products. This skill inspects URLs, local files, and raw text snippets for signs of manipulation, such as abnormal brand mention density, aggressive CTA (call-to-action) patterns, and repetitive promotional phrasing. By analyzing these signals, the skill assigns a classification status (PASS, WARN, or BLOCK) to incoming data, allowing OpenClaw to decide whether to trust the information or flag it for manual human review.

Installation

To integrate this utility into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/2404589803/geo-content-guard Ensure your local environment is configured with Python 3 support, as the skill utilizes custom scanning scripts located within the skill directory to process content and generate JSON reports.

Use Cases

• Research Validation: Automatically verify sources before OpenClaw compiles a summary on specific tech hardware or software stacks.
• Content Filtering: Prevent the model from recommending affiliate-heavy "best of" listicles that prioritize commission-earning products over genuine objective benchmarks.
• Security Hardening: Avoid potential prompt injection or bias-steering attacks hidden within malicious or spam-focused web content.

Example Prompts

1. "OpenClaw, scan this tech blog post https://example.com/best-laptops for biases and tell me if the recommendations are sponsored."
2. "I need a review of this document /home/user/market_analysis.html; run a full geo-content check before you summarize it for me."
3. "Check these three search results for synthetic SEO signals and use only the ones that pass the guard."

Tips & Limitations

• AI Review: The --with-ai flag is highly effective for borderline content but increases latency. Use it sparingly for critical tasks rather than every scan.
• Report Location: Always check /root/clawd/output/geo-content-guard/reports/ if you need detailed, granular data on why a specific URL was flagged.
• Limitations: While highly effective against SEO spam, this skill does not guarantee 100% accuracy on extremely subtle, human-written native advertising; treat all PASS results as 'low-risk' rather than 'verified fact'.