inno-reference-audit

This skill provides reference guidance for citation verification in academic writing. Use when the user asks about "citation verification best practices", "how to verify references", "preventing fake citations", or needs guidance on citation accuracy. This skill supports ml-paper-writing by providing detailed verification principles and common error patterns.

Quality

61%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/inno-reference-audit/SKILL.md

Quality

Content

52%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 content gives a clear, actionable verification workflow with concrete query examples and a worked example, but it is hurt by significant redundancy across sections and by ignoring its own bundle files. The reference and script bundles are described in their READMEs but never linked from SKILL.md, so a reader gets a monolithic guide instead of an overview pointing to deeper material.

Suggestions

Deduplicate the repeated failure-handling and 'Core Principle' sections; keep one authoritative version and remove the duplicates in Summary/Best Practices.

Add explicit links from SKILL.md to the bundle files (e.g. 'See references/common-errors.md for error patterns' and 'See references/verification-rules.md for matching rules') so the body acts as an overview.

Trim explanatory padding about why citations matter (paper rejection/retraction, reviewer perception) that Claude already knows, to improve token efficiency.

Dimension	Reasoning	Score
Conciseness	The body is noticeably verbose with redundant sections: the 'Core Principle' is stated at the top and again in Summary, 'Handling Verification Failures' appears twice (under Usage Guide and Best Practices) with overlapping steps, and 'Preventing Fake Citations' overlaps 'Common Pitfalls'; this matches the 2 anchor of several unnecessary/padded sections rather than 3's 'some' tightening needed.	2 / 5
Actionability	Provides concrete, runnable guidance via specific WebSearch query strings (e.g. '"site:scholar.google.com [paper title] [first author]"') and a fully worked Transformer-paper example with expected results; not a 5 because the queries are template strings rather than copy-paste-ready commands, and not a 3 because the guidance is concrete and specific rather than pseudocode-only.	4 / 5
Workflow Clarity	The Verification Workflow gives a clear sequenced flow (find → verify existence → confirm details → get BibTeX → verify claim → add) with explicit failure-handling feedback loops ('Mark as pending', 'Notify the user'); not a 5 because in-flow validation gates are implicit ('Confirm the paper appears') rather than hard checkpoints, and clearly above 3 since checkpoints and error recovery are present.	4 / 5
Progressive Disclosure	Bundle files exist (references/common-errors.md, verification-rules.md, api-usage.md and scripts/*.py) but the SKILL.md body never references or links to any of them, inlining summaries of that same material instead; this matches the 2 anchor (content that belongs in separate files is inlined, references buried) and is worse than 3 (which requires references present but unclearly signaled) because the references are entirely absent from the body.	2 / 5
	Total	12 / 20 Passed

Description

70%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 clearly answers both what the skill does and when to trigger it, with natural user phrasings as triggers. Its main weakness is that the 'what' is stated abstractly ('provides reference guidance') rather than as concrete operations, and it has minor overlap with the ml-paper-writing skill it supports.

Suggestions

Replace the abstract 'provides reference guidance' with concrete actions, e.g. 'Verifies citations exist on Google Scholar, matches author/title/year metadata, and flags fabricated references'.

Add common synonym triggers users might say, such as 'fake references', 'hallucinated citations', or 'check my bibliography'.

Clarify the boundary with ml-paper-writing so the trigger logic makes clear when this standalone skill should fire instead of the parent skill.

Dimension	Reasoning	Score
Specificity	Names the domain (citation verification in academic writing) and a couple of actions ('provides reference guidance', 'providing detailed verification principles and common error patterns'), but these are abstract rather than a comprehensive list of concrete actions, matching the 3 anchor; not a 4 because it lacks several specific operations, and not a 2 because it does name more than minimal actions.	3 / 5
Completeness	Explicitly states both what it does ('provides reference guidance for citation verification') and when to use it ('Use when the user asks about...') with concrete triggers; not a 5 because the 'what' is somewhat generic ('provides reference guidance') rather than listing concrete actions, and clearly above 3 since both what and when are present and explicit.	4 / 5
Trigger Term Quality	Includes several natural user phrasings in quotes ('citation verification best practices', 'how to verify references', 'preventing fake citations', 'needs guidance on citation accuracy') giving good keyword coverage; not a 5 because it omits common synonyms/variations a user might say, and not a 3 because the terms present are genuinely natural rather than generic.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (citation verification) with distinct trigger phrases and low conflict risk; not a 5 because it explicitly overlaps with and supports the ml-paper-writing skill, creating minor overlap risk, and not a 3 because the focus is specific rather than broad.	4 / 5
	Total	15 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

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

	Total	15 / 16 Passed

Repository: OpenLAIR/dr-claw
Path: skills/inno-reference-audit/SKILL.md
Commit: 4b35756

Reviewed: 5 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.