CtrlK
BlogDocsLog inGet started
Tessl Logo

rag-engineer

I bridge the gap between raw documents and LLM understanding. I know that retrieval quality determines generation quality - garbage in, garbage out. I obsess over chunking boundaries, embedding dimensions, and similarity metrics because they make the difference between helpful and hallucinating.

37

Quality

35%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/AI-Agents-Safe-Coding-Skills-claude/skills/rag-engineer/SKILL.md

The canonical home for this skill is rag-engineer in administrakt0r/AI-Agents-Safe-Coding-Skills

SKILL.md
Quality
Evals
Security

Quality

Content

40%Weight 40%Scale 1-5

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

The body is a well-sectioned but shallow overview: it lists capabilities and high-level pattern advice, yet provides no executable code, no sequenced workflow with validation, and several empty or filler sections. The ```javascript fences wrapping prose bullets and truncated table solutions are concrete defects.

Suggestions

Replace the misformatted ```javascript prose blocks with actual executable code (e.g. a real chunking snippet or a reciprocal-rank-fusion example) or convert them to plain bullet lists.

Complete the Sharp Edges table solutions and the empty Anti-Pattern sections, or remove them if not actionable.

Add a sequenced retrieval-build workflow with a validation checkpoint (e.g. evaluate retrieval quality separately from generation before shipping).

DimensionReasoningScore

Conciseness

The body is mostly lean lists, but it wastes tokens on concepts Claude already knows ('Vector embeddings', 'LLM fundamentals', 'Basic NLP concepts') and includes a content-free 'When to Use' sentence ('This skill is applicable to execute the workflow or actions described in the overview') and empty Anti-Pattern headers.

3 / 5

Actionability

The Patterns sections are prose advice bullets ('Use sentence boundaries, not token limits', 'Detect topic shifts with embedding similarity') mis-wrapped in ```javascript code fences rather than executable code, and the Sharp Edges table solutions are truncated to a colon with no actual fix, leaving only minimal concrete guidance.

2 / 5

Workflow Clarity

There is no sequenced workflow — Patterns are unordered bullet lists with no steps, and operations that touch indexing/embedding refreshes lack any validation or checkpoints, so the sequence is rough and validation is absent.

2 / 5

Progressive Disclosure

No bundle files exist and none are needed; the body is self-contained with clear section headers (Capabilities, Patterns, Anti-Patterns, Sharp Edges, Related Skills), so structure is good with only minor gaps like empty Anti-Pattern headers.

4 / 5

Total

11

/

20

Passed

Description

31%Weight 40%Scale 1-5

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

The description is a first-person, metaphor-heavy statement of philosophy rather than a concrete capability+trigger summary. It names the retrieval/RAG domain but specifies no executable actions and provides no 'Use when' guidance, capping completeness and specificity low.

Suggestions

Rewrite in third person with concrete verbs, e.g. 'Chunks documents, builds vector indexes, and tunes retrieval pipelines for LLM grounding.'

Add an explicit trigger clause: 'Use when building or debugging RAG systems, semantic search, or document-grounded LLM answers.'

Replace 'I obsess over...' framing with listed capabilities and natural user keywords like 'retrieval-augmented generation', 'semantic search', 'vector database'.

DimensionReasoningScore

Specificity

The description names the domain via aspirational framing ('I bridge the gap', 'I obsess over chunking boundaries, embedding dimensions, and similarity metrics') but lists no concrete invokable actions; first-person voice ('I bridge', 'I know', 'I obsess') triggers a 1-point specificity penalty, dropping it from 2 to 1.

1 / 5

Completeness

It gives only a metaphorical 'what' ('bridge the gap between raw documents and LLM understanding') with no concrete 'Use when...' trigger clause, so it falls between a vague 'what' with no 'when' (anchor 2) and a clear 'what' with missing 'when' (anchor 3); the missing trigger guidance caps it low.

2 / 5

Trigger Term Quality

It surfaces relevant jargon ('RAG'-adjacent terms: documents, LLM, embeddings, chunking, similarity metrics) but misses the natural phrases users actually say ('build a RAG system', 'set up retrieval', 'semantic search') and any synonyms or file extensions.

3 / 5

Distinctiveness Conflict Risk

The retrieval/RAG niche is somewhat distinct, but the broad framing ('bridge the gap between raw documents and LLM understanding') overlaps with adjacent skills like prompt-engineer and ai-agents-architect, leaving moderate conflict risk.

3 / 5

Total

9

/

20

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

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
administrakt0r/AI-Agents-Safe-Coding-Skills
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.