auto-paper-improvement-loop

Autonomously improve a generated paper via GPT-6-Astra 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

77%

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

Quality

Content

63%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 highly actionable, well-sequenced workflow for a complex paper-improvement loop, with strong validation checkpoints. It is held back by verbosity (duplicated prompts, narrative justifications) and a monolithic structure that inlines spec content a separate reference file would carry better.

Suggestions

De-duplicate the Round 1 and Round 2 review prompts: define the prompt once and have Round 2 add only its 'fresh, zero-context' delta, saving ~40 lines.

Move the full edit-whitelist spec (schema, resolution rules, glob semantics, detector table, behavior subsections) into a references/edit-whitelist.md file and keep a short summary + link in SKILL.md.

Complete the Step 4.5 Python comparison logic (the normalize() function is defined but the main-vs-appendix comparison is left as a comment) or explicitly point to /proof-checker --restatement-check as the executable path.

Dimension	Reasoning	Score
Conciseness	Mostly operational but noticeably padded: the ~50-line Round 1 and Round 2 review prompts are near-duplicates, and repeated 'Empirical motivation' narrative paragraphs justify rather than instruct.	3 / 5
Actionability	Provides executable bash (latexmk, for-loops, pdfinfo), full MCP review prompts, and concrete regex detectors; the minor gap is the Step 4.5 Python block that defines normalize() but stubs the actual comparison as a comment.	4 / 5
Workflow Clarity	Clear Step 0–9 sequence with explicit validation checkpoints (Step 4.5 restatement regression, Step 8 format stop-criteria), but retry-on-failure feedback loops are implied rather than spelled out as fix→revalidate loops.	4 / 5
Progressive Disclosure	Shared references are clearly signaled one level deep, but no bundle files exist and substantial spec content (the ~100-line edit-whitelist schema/rules and the duplicated review prompts) is inlined in SKILL.md where it belongs in separate reference files.	3 / 5
	Total	14 / 20 Passed

Description

91%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 with concrete actions, explicit trigger phrases in two languages, and clear what/when coverage. Its only weakness is minor overlap risk with sibling review-loop skills.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ('review → implement fixes → recompile, for 2 rounds') with a specific model and round count, but the action set is narrow for a single-purpose skill so coverage is not maximal.	4 / 5
Completeness	Explicitly answers both what ('Autonomously improve a generated paper via GPT-6-Astra xhigh review → implement fixes → recompile, for 2 rounds') and when ('Use when user says...') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive natural trigger coverage across two languages ('改论文', 'improve paper', '论文润色循环', 'auto improve', 'iteratively polish a generated paper') including synonyms and phrasings users would actually say.	5 / 5
Distinctiveness Conflict Risk	Has a clear niche (generated-paper writing-quality improvement) with distinct triggers, but minor overlap risk with the related /auto-review-loop skill that also iterates via review.	4 / 5
	Total	18 / 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 (696 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: 0472e53

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.