CtrlK
BlogDocsLog inGet started
Tessl Logo

regex-vs-llm-structured-text

Decision framework for choosing between regex and LLM when parsing structured text — start with regex, add LLM only for low-confidence edge cases.

58

Quality

67%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Advisory

Suggest reviewing before use

Fix and improve this skill with Tessl

tessl review fix ./skills/regex-vs-llm-structured-text/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

70%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, actionable skill with a clear sequenced pipeline and good section organization, but it is slightly verbose with redundant 'when to use' sections and one incomplete code path (the LLM validator). No external references are needed and none are missing.

Suggestions

Complete the LLM validator: replace the elided '# Parse LLM response and return corrected item...' with actual response parsing so the edge-case step is fully executable.

Merge the overlapping 'When to Activate' and 'When to Use' sections into one to remove redundancy and tighten token efficiency.

DimensionReasoningScore

Conciseness

Mostly efficient with tight code blocks and a useful metrics table, but several sections ('When to Activate' vs 'When to Use') overlap and the prose restates the regex/LLM cost idea multiple times, so it could be tightened without losing substance.

2 / 3

Actionability

Provides substantial executable Python (regex parser, confidence scorer, hybrid pipeline) but the LLM validator returns an undefined `corrected_item` with the parsing of the response elided ('# Parse LLM response and return corrected item...'), leaving the edge-case step incomplete.

2 / 3

Workflow Clarity

The pipeline is clearly sequenced (regex → confidence score → LLM for flagged items only) with an explicit confidence threshold checkpoint and a decision-tree diagram; the feedback path (low-confidence → LLM validate → output) is explicit and copy-paste ready.

3 / 3

Progressive Disclosure

No bundle files exist and none are needed; the single-file skill is well-organized into clearly labeled sections (Decision Framework, Architecture, Implementation, Metrics, Best Practices, Anti-Patterns) with no nested references, which satisfies the simple-skill bar of 3.

3 / 3

Total

10

/

12

Passed

Description

65%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A specific, well-framed description with a clear niche and concrete strategy, but it lacks an explicit 'Use when...' trigger clause, which limits completeness and trigger-term breadth. Voice is correctly third person.

Suggestions

Append an explicit trigger clause, e.g. 'Use when parsing quizzes, forms, invoices, or documents with repeating patterns and deciding whether to use regex or an LLM.'

Add concrete trigger nouns users would actually say ('quiz questions', 'form fields', 'invoice/receipt extraction') to broaden trigger-term coverage.

DimensionReasoningScore

Specificity

Names concrete actions — 'choosing between regex and LLM when parsing structured text' with an explicit strategy ('start with regex, add LLM only for low-confidence edge cases') — multiple specific mechanics rather than a single vague verb.

3 / 3

Completeness

Clearly states what the skill does (decision framework for regex vs LLM) but lacks an explicit 'Use when...' trigger clause, so the 'when' half is only implied — this caps completeness at 2 per the rubric guideline.

2 / 3

Trigger Term Quality

Includes natural terms like 'regex', 'LLM', and 'parsing structured text' but omits common user variations ('parse quiz questions', 'extract form fields', 'invoice parsing', 'document extraction'); coverage is decent but not broad.

2 / 3

Distinctiveness Conflict Risk

The regex-vs-LLM cost/accuracy tradeoff niche is distinct and unlikely to fire for unrelated parsing skills; the specific framing carves a clear, narrow trigger space.

3 / 3

Total

10

/

12

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository
affaan-m/ECC
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.