paper revision guardian

    1

    Professional academic manuscript revision and technical audit for CS, robotics, and ML research papers.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    paper revision guardian

    Example session with this skill installed

    Review the Introduction section of my robotics paper for clarity and tone, keeping the original authorial voice. Ensure no defensive language is used. Here is the text: [Introduction Text]

    • Read your context and instructions
    • Compiled the paper revision guardian

    I have reviewed the Introduction section focusing on clarity, technical precision, and the removal of defensive prose. I have preserved your original terminology and contribution-led narrative.

    Part 1 [Revised English]

    Recent developments in embodied AI have highlighted the need for robust tactile sensing in unstructured environments. While previous approaches relied heavily on visual feedback, our proposed framework utilizes a high-resolution force-torque sensor array to achieve stable grasping. We acknowledge that our method requires significant computational overhead; however, the resulting increase in success rate justifies this cost. Our contribution is twofold: first, a novel sensor fusion algorithm, and second, an open-source dataset of varied grasping scenarios.

    Part 2 [Chinese Translation]

    具身智能的最新发展凸显了在非结构化环境中对强健触觉感知的需求。虽然以前的方法在很大程度上依赖于视觉反馈,但我们提出的框架利用高分辨率力矩传感器阵列来实现稳定抓取。我们承认我们的方法需要大量的计算开销;然而,由此产生的抓取成功率的提高证明了这一成本是合理的。我们的贡献包括两个方面:第一,一种新型的传感器融合算法;第二,一个包含各种抓取场景的开源数据集。

    Part 3 [Modification Log and Technical Audit]

    1. [B] 改进表达: 将 "It might be argued that..." 修改为直接陈述。删除了原文中过多的防御性修饰语,使贡献陈述更加直接明了。
    2. [B] 风格一致性: 保留了作者对 "unstructured environments" 的特定定义,未进行同义词替换,以维持专业语境的一致性。
    3. [检测通过]: 技术术语(Embodied AI, Force-torque sensor array)使用准确,符合计算机机器人领域的学术惯例。
    4. [风险提示]: 文中提到的 "significant computational overhead" 尚未关联具体的实验数据,建议在 Results 部分补充量化对比。

    Next steps

    • Review the modified claim regarding computational overhead to ensure it matches your data in Table 2.
    • Confirm if the specific sensor array model should be named in the first mention here or in the Methods section.
    • Proceed to the Methods section for similar voice-consistent editing.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    This skill addresses the critical pain point of generic AI rewriting that strips the technical nuance and original voice from academic papers. Designed specifically for computer science, robotics, and machine learning researchers, it provides professional-grade manuscript revision and reviewer response management. It avoids the common mistake of "academic fluff" by prioritizing scientific integrity and evidence-claim consistency over cosmetic paraphrasing.

    What it does

    • Voice-preserved revision improves English clarity while strictly maintaining the author's original technical terminology and sentence rhythm.
    • Anti-defensive editing identifies and removes unnecessary defensive prose and repetitive apologies without hiding genuine study limitations.
    • Visual asset audit checks scientific figures, tables, and plots for consistency in color, typography, and layout while flagging watermark risks.
    • Mathematical verification reviews symbols, dimensions, constraints, and pseudocode logic within LaTeX environments.
    • Reviewer response mapping traces every reviewer concern to specific evidence, source locations, and revised manuscript sections.

    How it works

    1. Inventory phase: The agent catalogs manuscript sources, figures, datasets, and reviewer comments to define the audit scope.
    2. Style baseline: It reads unchanged prose to characterize the specific authorial voice before making any edits.
    3. Triaged classification: Findings are categorized as demonstrable errors, necessary editorial improvements, or unverified scientific claims.
    4. Targeted revision: Edits are applied to authorized regions only, providing a three-part delivery: revised English, Chinese translation, and a technical audit log.

    Frameworks & tools

    This skill is optimized for LaTeX environments and academic publishing standards like Advanced Robotics. It includes an optional Python-based tex_audit.py script for preliminary LaTeX screening.

    Why this beats prompting it yourself

    Generic prompts often result in "AI-flavored" academic prose that reviewers find suspicious or overly flowery. This skill uses a strict "minimal edit" policy and multi-step reference checks to ensure revisions actually solve clarity issues without manufacturing results or losing the researcher's unique perspective.

    Use cases

    • Refining a CS manuscript for journal submission while preserving specific algorithmic terminology.
    • Drafting point-by-point responses to peer reviewer critiques with verified evidence mapping.
    • Auditing experimental plots and architecture diagrams for visual style consistency.
    • Cleaning up defensive language in a Discussion section without compromising scientific honesty.

    Known limitations

    Does not invent experimental results or citations. Cannot certify mathematical correctness or guarantee journal acceptance. Requires access to source figures for visual audits.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Technical Documentation.

    Cannot invent data or citations. Does not guarantee journal acceptance. Preliminary LaTeX screening requires a Python runtime. Visual audits are limited to accessible figure files.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 21 days ago

    • Free with an Agensi account

    Software
    Git

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listed21 days ago

    What's inside

    Frequently Asked Questions

    Unlike generic AI that often introduces flowery 'academic fluff,' this skill uses a 'minimal edit' policy to fix English clarity while strictly preserving your specific technical terminology and original authorial voice.

    The skill is optimized for LaTeX environments and includes a specialized Python-based `tex_audit.py` script to screen your source files for structural and mathematical consistency.

    You will receive the professionally revised manuscript, a translation for context, a technical audit log of all changes, and a point-by-point response map for peer reviewer concerns.

    No, this skill is designed for integrity; it audits existing evidence-claim consistency and cleans up prose, but it will never invent citations or fabricate experimental data.

    This skill is compatible with leading AI agent platforms that support file uploads and Python script execution, particularly those used by CS and robotics researchers.

    Buyers receive the core agent skill, the LaTeX screening script, and access to all subsequent version updates that reflect changing journal standards like Advanced Robotics.

    New in Technical Documentation

    Free