source-selection-auditor

Evaluates whether an AI retriever can reliably select a page's primary factual content by measuring content density ratio, heading-to-answer locality, and…

Skill by AAX10

IndexedMIT

Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/aax10-source-selection-auditor

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

About

Evaluates whether an AI retriever can reliably select a page's primary factual content by measuring content density ratio, heading-to-answer locality, and proposition fragmentation.

Use cases: writing

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools: No tools found
Secrets: None asked for
Install scripts: None found
Runs as: Python package

Network hosts it talks to 1

codeload.github.com · found in install command

Files it touches

~/.claude/skills/source-selection-auditor

What's inside

Skill name: source-selection-auditor
Description: Evaluates whether an AI retriever can reliably select a page's primary factual content by measuring content density ratio, heading-to-answer locality, and proposition fragmentation.
License: MIT
Instructions: 22 lines in SKILL.md · view on GitHub

Before installing

Ships executable scripts

Files (4)

SKILL.md
density_heuristics.md · references/
evaluate_source_selection.py · scripts/
utils_html.py · scripts/

SKILL.md (first lines)

# Source Selection Auditor

## When to use
Use this skill when auditing a webpage to determine whether AI retrievers and RAG chunkers can extract the core factual content without being polluted by layout chrome, heading-answer disconnects, or cross-boundary proposition splits.

## Inputs
- `url` or local `.html` filepath: The target page to audit.

## Procedure
1. Execute `python scripts/evaluate_source_selection.py <url_or_filepath>`
2. The script extracts the DOM, detects the main content region via semantic landmarks.
3. Evaluates three core hypotheses:
   - **H-SS1**: Content Density Ratio (MCDR)
   - **H-SS2**: Heading-to-Answer Semantic Locality (HAL)
   - **H-SS3**: Fragmented Proposition Split
4. Computes composite Score_SS = 0.55 × MCDR + 0.45 × HAL
5. Generates proactive recommendations even on PASS pages.

## Output
A JSON report matching the Adobe Hackathon audit schema with `audited_at`, `site`, `score_ss`, `summary`, and `findings`.

From AAX10/Adobe-Round-3, MIT.

Signals

Stars: 0
Forks: 0
Open issues: 0
Last commit (this folder): Sep 11, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License: MIT
Repository: github.com/AAX10/Adobe-Round-3
First seen: Oct 24, 2026
Last checked: Oct 24, 2026
Source: GitHub github-graphql

More from AAX10

Cite this page

source-selection-auditor skill. Agent Mart. https://agentmart.dev/skills/aax10-source-selection-auditor (last updated 2026-10-24)