paper-claim-audit

Zero-context verification that every number, comparison, and scope claim in the paper matches raw result files. Uses a fresh cross-model reviewer with NO prior context to prevent confirmation bias. Use when user says "审查论文数据", "check paper claims", "verify numbers", "论文数字核对", or before submission to ensure paper-to-evidence fidelity.

Quality

88%

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

77%

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

A highly actionable, well-sequenced audit skill with strong validation checkpoints, but it is monolithic and verbose for its size. Splitting reference-heavy sections into bundle files and de-duplicating the fresh-thread invariant would improve token efficiency and navigation.

Suggestions

State the 'fresh thread, never codex-reply, zero executor interpretation' invariant once in Key Rules and reference it from Step 2 / Thread independence instead of repeating it verbatim in four places.

Move the full PAPER_CLAIM_AUDIT.json schema, the render-HTML section, and the Submission Artifact Emission details into reference files under references/ and link to them from SKILL.md to reduce inline bulk.

Trim the 'Why This Exists' confirmation-bias examples to one line; Claude already understands confirmation bias, so the four worked bullets add tokens without adding guidance.

Dimension	Reasoning	Score
Conciseness	The ~340-line body is mostly efficient but repeats the 'fresh thread / never codex-reply' invariant across Step 2, Key Rules, and the Thread independence section, and the 'Why This Exists' block explains confirmation bias with worked examples Claude already understands.	2 / 3
Actionability	Provides a concrete mcp__codex__codex call with model, reasoning-effort config, and a full prompt template, plus exact file-path patterns, MD+JSON output templates, and a complete artifact schema — all copy-paste ready.	3 / 3
Workflow Clarity	A clear 4-step sequence (Collect → Audit → Report → Summary) anchored by an explicit verdict decision table, a status enum, and PASS/WARN/FAIL advisory handling that act as validation checkpoints.	3 / 3
Progressive Disclosure	No bundle files exist and the skill is a single ~340-line document; the JSON schema, render-html, and submission-artifact-emission sections are inline where separate reference files would aid navigation, though shared-references links are well-signaled and one-level deep.	2 / 3
	Total	10 / 12 Passed

Description

100%

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 high-quality description: specific, third-person, with explicit bilingual triggers and a clear what/when structure. It is distinct from sibling audit skills and avoids fluff.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'verifies that every number, comparison, and scope claim in the paper matches raw result files' and 'Uses a fresh cross-model reviewer with NO prior context' — rather than vague language.	3 / 3
Completeness	Clearly answers both what (zero-context verification of numbers/comparisons/scope claims against raw files) and when, with an explicit 'Use when user says…' trigger clause.	3 / 3
Trigger Term Quality	Good coverage of natural terms users would say, including bilingual triggers: 'check paper claims', 'verify numbers', '审查论文数据', '论文数字核对', plus 'before submission'.	3 / 3
Distinctiveness Conflict Risk	Occupies a clear niche (paper-to-evidence fidelity audit) with distinct triggers; unlikely to fire for experiment-audit or result-to-claim skills.	3 / 3
	Total	12 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 suspicious	Warning

	Total	13 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/paper-claim-audit/SKILL.md
Commit: c5f3d5b

Reviewed: about 1 hour 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.