regex-vs-llm-structured-text — Claude Code skill: parsing text
The regex-vs-llm-structured-text Claude Code skill helps choose an extraction method for structured text. It starts with regex for repeated patterns and sends uncertain results to an LLM for review.
MIT 6,200 installs ★ 308,413 updated 2026-10-20
npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a claude-codeWhat regex-vs-llm-structured-text does
The skill lays out a hybrid parsing flow: regex extracts fields, a confidence score flags questionable results, and an LLM checks only those items. The examples cover quizzes, forms, invoices, receipts, and tables with recurring layouts.
It includes a Python parser, confidence-scoring functions, and a validation pipeline. For free-form text with highly variable structure, the guidance switches to using an LLM directly. Its validation example calls Claude Haiku through a messages client; LLM requests are reserved for items below the confidence threshold.
What it helps with
- extract questions and answers from repeated text blocks
- parse fields from forms, invoices, or receipts with regex
- send only uncertain extractions to an LLM for review
- choose regex or an LLM for text with variable structure
Example prompts for the agent
- Show me a regex approach for extracting questions and answers from repeated text.
- Add confidence scoring and send questionable extracted items to an LLM for review.
- Compare regex and LLM parsing for a form whose layout varies.
How to install regex-vs-llm-structured-text
1Skills CLI (Vercel)
npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a claude-code2Manually: put the files in place
~/.claude/skills/regex-vs-llm-structured-text/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/regex-vs-llm-structured-text/
1Skills CLI (Vercel)
npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a codex2Manually: put the files in place
.agents/skills/regex-vs-llm-structured-text/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a cursor2Manually: put the files in place
.cursor/skills/regex-vs-llm-structured-text/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install affaan-m/ECC skills/regex-vs-llm-structured-text2Manually: put the files in place
.github/skills/regex-vs-llm-structured-text/SKILL.md1Skills CLI (Vercel)
npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a gemini-cli2Platform CLI
gemini skills install https://github.com/affaan-m/ECC --path skills/regex-vs-llm-structured-textManually: put the files in place
.devin/skills/regex-vs-llm-structured-text/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about regex-vs-llm-structured-text
How do I install regex-vs-llm-structured-text in Claude Code?
Run npx skills add affaan-m/ECC --skill regex-vs-llm-structured-text -a claude-code. The skill file is at ~/.claude/skills/regex-vs-llm-structured-text/SKILL.md.
What does regex-vs-llm-structured-text do?
It lays out a parsing approach that uses regex for recurring structures, scores extraction confidence, and sends low-confidence items to an LLM for validation.
When does the skill recommend using an LLM instead of regex?
For free-form text with highly variable structure, it recommends using an LLM directly. Regex is the suggested starting point for repeated formats.
Source text
SKILL.md · MIT
---
name: regex-vs-llm-structured-text
description: Decision framework for parsing structured text (quizzes, forms, invoices, receipts, tables) with a hybrid regex-first pipeline — regex extraction handles 95%+ cheaply, a confidence scorer flags low-confidence items, and an LLM validator fixes only the edge cases. Use when choosing between regex and LLM for text extraction, building a cheap document parser, or optimizing extraction cost and accuracy.
metadata:
origin: ECC
---
# Regex vs LLM for Structured Text Parsing
A practical decision framework for parsing structured text (quizzes, forms, invoices, documents). The key insight: regex handles 95-98% of cases cheaply and deterministically. Reserve expensive LLM calls for the remaining edge cases.
## When to Activate
- Parsing structured text with repeating patterns (questions, forms, tables)
- Deciding between regex and LLM for text extraction
- Building hybrid pipelines that combine both approaches
- Optimizing cost/accuracy tradeoffs in text processing
## Decision Framework
```
Is the text format consistent and repeating?
├── Yes (>90% follows a pattern) → Start with Regex
│ ├── Regex handles 95%+ → Done, no LLM needed
│ └── Regex handles <95% → Add LLM for edge cases only
└── No (free-form, highly variable) → Use LLM directly
```
## Architecture Pattern
```
Source Text
│
▼
[Regex Parser] ─── Extracts structure (95-98% accuracy)
│
▼
[Text Cleaner] ─── Removes noise (markers, page numbers, artifacts)
│
▼
[Confidence Scorer] ─── Flags low-confidence extractions
│
├── High confidence (≥0.95) → Direct output
│
└── Low confidence (<0.95) → [LLM Validator] → Output
```
## Implementation
### 1. Regex Parser (Handles the Majority)
```python
import re
from dataclasses import dataclass
@dataclass(frozen=True)
class ParsedItem:
id: str
text: str
choices: tuple[str, ...]
answer: str
confidence: float = 1.0
def parse_structured_text(content: str) -> list[ParsedItem]:
"""Parse structured text using regex patterns."""
pattern = re.compile(
r"(?P<id>\d+)\.\s*(?P<text>.+?)\n"
r"(?P<choices>(?:[A-D]\..+?\n)+)"
r"Answer:\s*(?P<answer>[A-D])",
re.MULTILINE | re.DOTALL,
)
items = []
for match in pattern.finditer(content):
choices = tuple(
c.strip() for c in re.findall(r"[A-D]\.\s*(.+)", match.group("choices"))
)
items.append(ParsedItem(
id=match.group("id"),
text=match.group("text").strip(),
choices=choices,
answer=match.group("answer"),
))
return items
```
### 2. Confidence Scoring
Flag items that may need LLM review:
```python
@dataclass(frozen=True)
class ConfidenceFlag:
item_id: str
score: float
reasons: tuple[str, ...]
def score_confidence(item: ParsedItem) -> ConfidenceFlag:
"""Score extraction confidence and flag issues."""
reasons = []
score = 1.0
if len(item.choices) < 3:
reasons.append("few_choices")
score -= 0.3
if not item.answer:
reasons.append("missing_answer")
score -= 0.5
if len(item.text) < 10:
reasons.append("short_text")
score -= 0.2
return ConfidenceFlag(
item_id=item.id,
score=max(0.0, score),
reasons=tuple(reasons),
)
def identify_low_confidence(
items: list[ParsedItem],
threshold: float = 0.95,
) -> list[ConfidenceFlag]:
"""Return items below confidence threshold."""
flags = [score_confidence(item) for item in items]
return [f for f in flags if f.score < threshold]
```
### 3. LLM Validator (Edge Cases Only)
```python
def validate_with_llm(
item: ParsedItem,
original_text: str,
client,
) -> ParsedItem:
"""Use LLM to fix low-confidence extractions."""
response = client.messages.create(
model="claude-haiku-4-5-20251001", # Cheapest model for validation
max_tokens=500,
messages=[{
"role": "user",
"content": (
f"Extract the question, choices, and answer from this text.\n\n"
f"Text: {original_text}\n\n"
f"Current extraction: {item}\n\n"
f"Return corrected JSON if needed, or 'CORRECT' if accurate."
),
}],
)
# Parse LLM response and return corrected item...
return corrected_item
```
### 4. Hybrid Pipeline
```python
def process_document(
content: str,
*,
llm_client=None,
confidence_threshold: float = 0.95,
) -> list[ParsedItem]:
"""Full pipeline: regex -> confidence check -> LLM for edge cases."""
# Step 1: Regex extraction (handles 95-98%)
items = parse_structured_text(content)
# Step 2: Confidence scoring
low_confidence = identify_low_confidence(items, confidence_threshold)
if not low_confidence or llm_client is None:
return items
# Step 3: LLM validation (only for flagged items)
low_conf_ids = {f.item_id for f in low_confidence}
result = []
for item in items:
if item.id in low_conf_ids:
result.append(validate_with_llm(item, content, llm_client))
else:
result.append(item)
return result
```
## Real-World Metrics
From a production quiz parsing pipeline (410 items):
| Metric | Value |
|--------|-------|
| Regex success rate | 98.0% |
| Low confidence items | 8 (2.0%) |
| LLM calls needed | ~5 |
| Cost savings vs all-LLM | ~95% |
| Test coverage | 93% |
## Best Practices
- **Start with regex** — even imperfect regex gives you a baseline to improve
- **Use confidence scoring** to programmatically identify what needs LLM help
- **Use the cheapest LLM** for validation (Haiku-class models are sufficient)
- **Never mutate** parsed items — return new instances from cleaning/validation steps
- **TDD works well** for parsers — write tests for known patterns first, then edge cases
- **Log metrics** (regex success rate, LLM call count) to track pipeline health
## Anti-Patterns to Avoid
- Sending all text to an LLM when regex handles 95%+ of cases (expensive and slow)
- Using regex for free-form, highly variable text (LLM is better here)
- Skipping confidence scoring and hoping regex "just works"
- Mutating parsed objects during cleaning/validation steps
- Not testing edge cases (malformed input, missing fields, encoding issues)
## When to Use
- Quiz/exam question parsing
- Form data extraction
- Invoice/receipt processing
- Document structure parsing (headers, sections, tables)
- Any structured text with repeating patterns where cost matters
© affaan-m · MIT · affaan-m/ECC