autorag

Search, summarize, compare, and answer questions from an already configured AutoRAG librarian over local documents and authorized datasources. Use when the user asks AutoRAG to search PDFs, wikis, notes, or a knowledge base. Use autorag-setup for install, model, roots, indexing, or datasource changes.

Quality

81%

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

Quality

Content

76%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.

A lean, highly actionable CLI skill body with copy-paste commands and good section structure. The main gap is the destructive maintenance commands lacking explicit validate-before-proceed checkpoints, which caps workflow clarity.

Suggestions

Add explicit validation steps around destructive maintenance (e.g. run `autorag health --json` before `autorag index rebuild`/`reset`, and confirm before `refresh --force`) to create a validate -> act -> re-verify feedback loop.

Move the long Maintenance command bank into a short "common commands" table or a referenced file so the overview stays scannable, raising progressive disclosure toward 5.

Tighten the Preflight prose (e.g. the duplicate-file and MinSync/Jikji paragraphs) to trim the few explanatory passages and push conciseness to 5.

Dimension	Reasoning	Score
Conciseness	The body is dense and operational, assuming Claude's competence (e.g. "BM25 is a MinSync retrieval mode, not a --method name"), with only minor explanatory passages that could be trimmed, fitting the efficient-but-slightly-padded anchor.	4 / 5
Actionability	Copy-paste-ready commands cover the common cases (search, evidence, feedback) plus a maintenance command bank, all fully executable with concrete flags and example arguments.	5 / 5
Workflow Clarity	The preflight and search flows sequence health checks and evidence inspection, but the destructive maintenance operations (index reset, rebuild, refresh --force) are listed without validation/feedback loops; per the rubric's destructive-operations cap this holds at 3 rather than 4.	3 / 5
Progressive Disclosure	Clear section structure (Preflight, Search, Maintenance, Rules) with a single well-signaled one-level reference (docs/embedding-runtime.md); most content is appropriately inline for a CLI skill, with only minor organization gaps.	4 / 5
	Total	16 / 20 Passed

Description

87%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.

A strong, third-person description that answers what and when with concrete triggers and a clear boundary against the sibling setup skill. Specificity is solid but the action verbs are slightly more abstract than a best-in-class enumerated list.

Dimension	Reasoning	Score
Specificity	Names four concrete actions ("Search, summarize, compare, and answer questions") over a defined domain (AutoRAG librarian over local documents and authorized datasources), but the verbs stay somewhat abstract compared to the level-5 example's granular action list.	4 / 5
Completeness	Explicitly states both what ("Search, summarize, compare, and answer questions from an already configured AutoRAG librarian") and when ("Use when the user asks AutoRAG to search PDFs, wikis, notes, or a knowledge base") with concrete trigger phrases, plus a boundary pointer to autorag-setup.	5 / 5
Trigger Term Quality	Natural triggers like "search PDFs, wikis, notes, or a knowledge base" are present and user-facing, but coverage lacks common synonyms/extensions (e.g. .pdf, research papers) needed for a 5.	4 / 5
Distinctiveness Conflict Risk	The named AutoRAG niche and explicit handoff to autorag-setup for install/indexing/datasource work give it a clear scope with minimal overlap risk against other skills.	5 / 5
	Total	18 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: Marker-Inc-Korea/AutoRAG
Path: skills/autorag/SKILL.md
Commit: fd4441a

Reviewed: 2 hours 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.