auto-paper-improvement-loop

Autonomously improve a generated paper via GPT-5.6-Sol xhigh review → implement fixes → recompile, for 2 rounds. Use when user says "改论文", "improve paper", "论文润色循环", "auto improve", or wants to iteratively polish a generated paper.

Quality

78%

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/auto-paper-improvement-loop/SKILL.md

Quality

Content

62%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 is highly actionable and has a clearly sequenced, validation-rich workflow, but it is markedly verbose and monolithic rather than progressively disclosed across files. Tightening prose and splitting reference material into separate files would raise the weaker dimensions.

Suggestions

Move the edit-whitelist schema/resolution rules, the fix-pattern tables, and the full reviewer prompt templates into separate reference files referenced one level deep, keeping SKILL.md as a lean overview.

Trim the repeated 'Empirical motivation' rationale blocks and conceptual exposition to reduce padding; assume Claude's competence about LaTeX compilation and review conventions.

Consolidate the duplicated Round 1 / Round 2 review prompt and fix-implementation sections via a shared template reference instead of restating each round in full.

Dimension	Reasoning	Score
Conciseness	At ~650 lines the body is noticeably verbose, with repeated rationale blocks, empirical-motivation prose, and large fix-pattern tables that assume context Claude already has; a tighter version could remove substantial padding.	2 / 5
Actionability	Provides mostly executable guidance with concrete bash/python snippets and ready-to-use review prompts, with only minor gaps (e.g., [VENUE]/[list figure files] placeholders left for the runner).	4 / 5
Workflow Clarity	Steps 0–9 are clearly sequenced with explicit validation checkpoints (compile verification, Step 4.5 restatement regression, Step 8 format check with stop criteria) and feedback loops for error recovery.	5 / 5
Progressive Disclosure	Content has section headers but is a monolithic single-file body with no bundle files; large reference-style material (whitelist schema, fix-pattern tables, reviewer prompts) is inlined rather than split into one-level-deep files.	3 / 5
	Total	14 / 20 Passed

Description

95%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 strong: it states concrete actions, provides comprehensive bilingual trigger terms, and explicitly covers both what and when. It is concise and distinct from related skills.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ("review → implement fixes → recompile, for 2 rounds") with only minor coverage gaps, sitting just below the comprehensive multi-action anchor.	4 / 5
Completeness	Explicitly answers both what it does (improve a paper via review→fix→recompile for 2 rounds) and when to use it with a concrete "Use when..." clause and trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive natural triggers including synonyms and bilingual variants ("改论文", "improve paper", "论文润色循环", "auto improve", "iteratively polish a generated paper") that users would actually say.	5 / 5
Distinctiveness Conflict Risk	Clear niche (paper-writing polish loop) with distinct triggers and explicit contrast to /auto-review-loop, giving minimal conflict risk.	5 / 5
	Total	19 / 20 Passed

Validation

75%

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 — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (652 lines); consider splitting into references/ and linking	Warning
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	12 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/auto-paper-improvement-loop/SKILL.md
Commit: f4f20f9

Reviewed: 1 day 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.