Academic Writer

PassAudited by ClawScan on May 1, 2026.

Overview

The skill is coherent for LaTeX writing and PDF compilation, with expected local file, shell, dependency, and cross-skill citation workflow risks that users should manage carefully.

This appears suitable for its stated academic-writing purpose, but install it only if you are comfortable with a skill that can read and write local project files, run LaTeX compilation, and use another research skill for citations. Keep backups, use a dedicated project directory, review setup commands before running them, and avoid providing sensitive unpublished material unless you trust the full workflow.

Findings (5)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

A mistaken or overly broad path could read unintended local text files or overwrite the wrong file.

Why it was flagged

The helper reads and writes caller-supplied paths, and write_latex can overwrite files by default. This is central to the writing workflow, but it is not constrained to a project directory.

Skill content
with open(file_path, 'r', encoding='utf-8', errors='ignore') as f: return f.read() ... with open(filename, mode, encoding='utf-8') as f: f.write(content)
Recommendation

Use explicit project-relative paths, avoid absolute or parent-directory paths unless intentional, and keep backups of important LaTeX files.

What this means

Compiling a project will run LaTeX tooling and create or update build artifacts in the working directory.

Why it was flagged

The skill invokes the local LaTeX toolchain through latexmk to compile PDFs. This is expected for the stated purpose, but it is still local command execution over project files.

Skill content
subprocess.run(['latexmk', '-pdf', '-interaction=nonstopmode', main_file], ... timeout=60)
Recommendation

Compile only trusted LaTeX projects/templates and review generated or modified files if the project came from an untrusted source.

What this means

Installing these dependencies can significantly change the local environment and relies on package-manager/PyPI provenance.

Why it was flagged

The setup instructions require manual installation of a large system LaTeX distribution and an unpinned Python package. This is disclosed and purpose-aligned, but not captured by the registry requirements.

Skill content
sudo apt-get install texlive-full ... pip install python-docx
Recommendation

Run setup commands manually after review, prefer a virtual environment for Python packages, and consider pinning or auditing dependencies if used in a sensitive environment.

What this means

Private drafts or notes may be used in the agent's working context, and inaccurate or instruction-like text inside those files could influence the generated writing.

Why it was flagged

The skill intentionally loads user-provided reference documents into the agent context and treats them as authoritative source material.

Skill content
If the user mentions input files ... call `read_reference`. ... Use this content as the "Ground Truth" for your writing.
Recommendation

Provide only documents you intend the agent to read, and review generated text for factual accuracy and unintended influence from source files.

What this means

Citation topics, claims, or related research context may be passed to another skill during writing.

Why it was flagged

The skill explicitly delegates citation retrieval to another skill. This is expected for the research workflow, but it introduces another component and data boundary.

Skill content
Designed to work in tandem with `academic-research-hub` for citation retrieval ... Calls `academic-research-hub`
Recommendation

Review the academic-research-hub skill before using it with unpublished or sensitive research topics.