rag-implementation

RAG (Retrieval-Augmented Generation) implementation workflow covering embedding selection, vector database setup, chunking strategies, and retrieval optimization.

Quality

33%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/rag-implementation/SKILL.md

Quality

Content

21%Scale 1-5

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

This skill is essentially a high-level project plan template that delegates all actual implementation to other skills via '@skill-name' references, providing no concrete, executable guidance itself. Every phase follows an identical template of vague action items and delegation prompts, resulting in significant verbosity with minimal actionable content. It reads more like a table of contents for a RAG curriculum than an implementable workflow.

Suggestions

Replace vague action items with concrete, executable code examples — e.g., show actual chunking code with LangChain/LlamaIndex, a real vector DB setup with connection strings, or a working retrieval pipeline snippet.

Add specific validation steps with concrete criteria — e.g., 'Retrieval accuracy should exceed 85% on test set; measure using MRR@10' instead of 'Evaluation passing'.

Consolidate the 8 repetitive phases into a more concise format — the identical 'Skills to Invoke / Actions / Copy-Paste Prompts' structure across all phases wastes tokens without adding value.

Include at least one end-to-end minimal working example (e.g., a 20-line Python script showing query → embed → retrieve → generate) so the skill provides standalone value beyond just referencing other skills.

Dimension	Reasoning	Score
Conciseness	The skill is heavily padded with repetitive structure across 8 phases that all follow the same template (Skills to Invoke, Actions, Copy-Paste Prompts). The numbered action items are vague platitudes ('Define use case', 'Select model') that Claude already knows. The 'When to Use This Workflow' section explains obvious use cases. Significant token waste with minimal unique information.	2 / 5
Actionability	The content is entirely abstract with no executable code, no concrete commands, no specific configurations, and no real examples. Every 'action' is a vague directive like 'Choose vector database' or 'Implement chunking'. The 'Copy-Paste Prompts' are just references to other skills (e.g., 'Use @rag-engineer to implement chunking strategy') with no actual implementation guidance.	1 / 5
Workflow Clarity	While there is a rough 8-phase sequence and a quality gates checklist, the steps within each phase are vaguely defined ('Implement vector search', 'Test retrieval quality') with no validation checkpoints, no error recovery, and no concrete criteria for moving between phases. The quality gates are just a checkbox list with no pass/fail criteria.	2 / 5
Progressive Disclosure	The skill references many external skills/bundles which suggests an intent for progressive disclosure, but no bundle files are provided to verify these references exist. The structure is organized into phases with clear headers, but the content is repetitive and could be significantly condensed. The architecture diagram is a nice touch but the overall organization is mediocre.	3 / 5
	Total	8 / 20 Passed

Description

45%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 identifies a clear domain (RAG implementation) and lists relevant sub-topics, but reads more like a course syllabus than an actionable skill description. It lacks concrete actions (what Claude will actually do) and entirely omits 'when to use' guidance, making it harder for Claude to reliably select this skill from a large pool. The technical terminology is relevant but could be expanded with more natural user phrases and synonyms.

Suggestions

Add an explicit 'Use when...' clause with trigger phrases like 'Use when the user asks about building a RAG pipeline, setting up semantic search, implementing document retrieval, or working with vector databases.'

Replace topic-area language with concrete actions, e.g., 'Guides implementation of RAG pipelines: selects embedding models, configures vector databases (Pinecone, ChromaDB, FAISS), implements chunking strategies, and optimizes retrieval quality.'

Include common synonyms and related terms users might naturally say, such as 'semantic search', 'knowledge base', 'document Q&A', 'similarity search', or specific vector DB names.

Dimension	Reasoning	Score
Specificity	Names the domain (RAG) and lists several sub-topics (embedding selection, vector database setup, chunking strategies, retrieval optimization), but these read more like topic areas than concrete actions. It doesn't specify what Claude actually does—e.g., 'generates embedding code', 'configures vector DB', 'implements chunking logic'.	3 / 5
Completeness	It provides a partial 'what' (covering RAG implementation topics) but has no 'when' clause at all. There is no 'Use when...' guidance, which per the rubric should cap completeness at 3, and since the 'what' is also somewhat vague (topic areas rather than concrete actions), a score of 2 is appropriate.	2 / 5
Trigger Term Quality	Includes relevant technical terms like 'RAG', 'Retrieval-Augmented Generation', 'embedding', 'vector database', 'chunking', and 'retrieval optimization'. However, it misses common natural user phrases like 'semantic search', 'knowledge base', 'document retrieval', 'similarity search', 'Pinecone', 'ChromaDB', 'FAISS', or other synonyms users might naturally use.	3 / 5
Distinctiveness Conflict Risk	RAG is a fairly specific niche, and the combination of embedding, vector database, chunking, and retrieval optimization creates a distinct profile. Minor overlap risk exists with general ML/AI skills or embedding-specific skills, but the RAG focus is clear enough to be mostly distinct.	4 / 5
	Total	12 / 20 Passed

Validation

90%

Warnings & errors only

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

Validation — 10 / 11 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	10 / 11 Passed

Repository: administrakt0r/AI-Agents-Safe-Coding-Skills
Path: skills/rag-implementation/SKILL.md
Commit: 57c135b

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.