Skip to content

[New Skill]: Agent-First Web-Proxy #42

@rosspeili

Description

@rosspeili

Skill Name

data_engineering/semantic_web_proxy

What should this skill do?

Standard Web HTML is incredibly "noisy" (full of styling, scripts, and ads), which wastes expensive LLM tokens and decreases reasoning accuracy. This skill acts as a proxy: it takes a raw URL, renders it, strips all non-semantic elements, and converts the content into highly concentrated, token-efficient Markdown or JSON that the LLM can easily digest.

Ideal Inputs & Outputs

Input:
{
"url": "https://news.ycombinator.com/item?id=12345",
"output_format": "markdown",
"focus_element": "main_content_and_comments"
}

Output:
{
"status": "success",
"token_count_saved": 14500,
"semantic_payload": "# Hacker News Thread...\n\n- User1: This is great."
}

Targeted Models (if applicable)

Model Agnostic (All)

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or requestskill requestRequest for a new capability to be added.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions