CtrlK
BlogDocsLog inGet started
Tessl Logo

auto-paper-improvement-loop

Autonomously improve a generated paper via Gemini review through gemini-review MCP → 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.

73

Quality

91%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

88%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 skill body with concrete commands, complete MCP call templates, and explicit validation feedback loops. The main improvement area is conciseness — the empirical score-progression anecdote and a duplicated polling paragraph could be trimmed.

Suggestions

Remove or relocate the "Typical Score Progression" anecdote ("Based on end-to-end testing on a real theory-paper run...") to keep the body lean and avoid time-sensitive empirical claims.

De-duplicate the MCP polling instructions that appear verbatim in both Step 2 and Step 5 — define the poll-until-done pattern once and reference it.

Consider moving the LaTeX fix-pattern tables (Steps 3 and 8) into a reference file to shorten the main body while keeping them one click away.

DimensionReasoningScore

Conciseness

Efficient and well-organized with concrete commands and tables rather than padding, but includes minor trimmable content such as the anecdotal "Typical Score Progression" narrative and a duplicated polling-instruction paragraph in Steps 2 and 5.

4 / 5

Actionability

Fully executable guidance throughout — concrete bash (latexmk, pdfinfo, grep), complete MCP tool-call blocks with YAML, review prompt templates, and fix-pattern tables with specific LaTeX remedies like \resizebox and \looseness=-1.

5 / 5

Workflow Clarity

A clearly sequenced Step 0–9 workflow with explicit validation checkpoints (verify 0 undefined references, overfull-hbox format check, "fix it and recompile before documenting") and feedback loops, plus explicit state-persistence recovery logic.

5 / 5

Progressive Disclosure

Good section structure with shared protocols referenced one level deep via ../../shared-references links; no bundle files exist to verify, and most content is appropriately inline with minor organization gaps in the longer sections.

4 / 5

Total

18

/

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.

A strong description: it states concrete actions, gives a precise 2-round scope, and provides rich multilingual trigger phrases for when to invoke it. The only minor gap is that the nature of the fixes is not enumerated at the description level.

DimensionReasoningScore

Specificity

Names the domain (paper improvement) and several concrete actions — "Gemini review through gemini-review MCP", "implement fixes", "recompile, for 2 rounds" — with only minor gaps in coverage of what kinds of fixes are applied.

4 / 5

Completeness

Explicitly answers both "what" (improve via Gemini review → implement fixes → recompile, 2 rounds) and "when" via a concrete "Use when user says..." clause with multiple trigger phrases.

5 / 5

Trigger Term Quality

Comprehensive natural trigger coverage including multilingual synonyms — "改论文", "improve paper", "论文润色循环", "auto improve", and the descriptive phrase "iteratively polish a generated paper".

5 / 5

Distinctiveness Conflict Risk

Clear niche (Gemini-reviewed iterative paper polishing) with distinct, specific triggers; minimal overlap risk with generic review skills.

5 / 5

Total

19

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 3 suspicious

Warning

Total

15

/

16

Passed

Repository
wanshuiyin/Auto-claude-code-research-in-sleep
Reviewed

Table of Contents

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.