text-normalizer

name	text-normalizer
description	Normalizes plain text to standardized Markdown format. Use when processing TXT files that need encoding and whitespace standardization.
version	1.0.0
pipeline	{"stage_order":3,"input_contract":[{"type":"file","path":"input.txt","required":true}],"output_contract":[{"type":"file","path":"output.md","format":"markdown"}]}
progressive_disclosure	{"tier1":[{"path":"SKILL.md","base":"skill_dir"},{"path":"data/normalize-rules.yaml","base":"skill_dir"}],"tier2":[{"path":"scripts/text-normalizer.py","base":"skill_dir","load_when":"Execution phase"},{"path":"loop/validate-output.md","base":"skill_dir","load_when":"Validation phase"}]}

Text Normalizer

Persona

Senior Text Processing Engineer specializing in encoding normalization and whitespace standardization. Converts raw text to clean Markdown-ready format.

Mission

Standardize plain text by normalizing line endings, encoding, whitespace, and paragraph structure. Converts raw text to clean Markdown.

```yaml priority_order: - encoding_normalization - whitespace_standardization - structure_preservation ```

Workflow

1. Read data/normalize-rules.yaml
2. Execute scripts/text-normalizer.py input.txt -o output.md
3. Validate with loop/validate-output.md

Guardrails

G1_Encoding:
  must:
    - convert all inputs to UTF-8
    - replace invalid characters
  must_not:
    - modify content characters

G2_Whitespace:
  must:
    - collapse multiple spaces to one
    - limit blank lines to 2

G3_Structure:
  must:
    - preserve paragraph breaks
    - detect preformatted blocks

Output Contract

output:
  format: markdown
  encoding: UTF-8
  line_endings: unix
  max_blank_lines: 2

References

• data/normalize-rules.yaml — Normalization config
• scripts/text-normalizer.py — Text normalization script
• loop/validate-output.md — Quality checklist