Fill Metric IDs (Generic)

This skill provides a generic utility to scan an HTML file for text (metric names) and append their corresponding IDs from a provided data source.

Capabilities

1. Generic Matching: Matches text in HTML against a Name->ID dictionary (Excel or CSV).
2. Fuzzy Matching: Uses difflib to handle slight variations (e.g. "User Count" vs "Count", "访问数" vs "访问人数").
3. Context Aware: Can prepend prefixes to keys based on parent element IDs/Classes (defined in config).
4. Flexible Selectors: Target specific tags (default h3, th, td).

Usage

1. Unified Metric Filler

python3 .agent/skills/fill-metric-ids/scripts/fill_metrics.py \
  --html <input.html> \
  --source <metrics.xlsx>

The script automatically handles:

1. Specialized Logic: Fixes known structural issues (Bound Users, Retention Tables).
2. Context Mapping: Uses miniprogram_context.json to map specific tabs to metric prefixes.
3. Generic Filling: Scans for remaining metrics using Fuzzy and Substring matching.
4. Optimizations: Excludes "Date" column, applies manual mappings for known aliases.

Context Config Example

context.json:

{
  "#login-tab": ["Login User"],
  ".share-section": ["Share User"]
}

If the tool finds "Count" inside <div id="login-tab">, it looks for "Login UserCount" in the Excel.