Excel Report

AdvisoryAudited by Static analysis on May 10, 2026.

Overview

Detected: suspicious.dynamic_code_execution

Findings (1)

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.

ConcernHigh Confidence

ASI05: Unexpected Code Execution

What this means

A malicious or tampered template could run code on the user's machine when report calculations are performed.

Why it was flagged

Formula text from template configuration is converted into a Python expression and evaluated without a sandbox or allowlist, so an untrusted or modified template formula could execute arbitrary Python code.

Skill content

result = eval(expr)

Recommendation

Replace eval with a restricted formula parser or AST-based evaluator that only permits expected arithmetic, column references, and approved aggregation functions.

NoteHigh Confidence

ASI03: Identity and Privilege Abuse

What this means

If enabled, the skill can use the configured email account to send generated reports.

Why it was flagged

The email feature requires SMTP account credentials and can send mail on the user's behalf; this is purpose-aligned but sensitive.

Skill content

export SMTP_HOST=smtp.gmail.com
export SMTP_USER=your@email.com
export SMTP_PASSWORD=xxx

python scripts/generate_report.py ... --email --to user@email.com

Recommendation

Use a limited app password or dedicated SMTP account, verify recipients, and avoid storing long-lived credentials in shared shell history or environments.

NoteHigh Confidence

ASI07: Insecure Inter-Agent Communication

What this means

Business, financial, medical, or operational data included in a report may leave the local environment via email.

Why it was flagged

Generated reports derived from local data can be sent through an external email provider when the user enables the email option.

Skill content

批量生成 + 邮件发送 ... --email --to <addr>

Recommendation

Only use email sending for intended recipients, review the generated workbook first, and avoid emailing sensitive reports unless the SMTP provider and recipient are trusted.

NoteMedium Confidence

ASI04: Agentic Supply Chain Vulnerabilities

What this means

Users may install whatever current versions are available from their configured Python package index.

Why it was flagged

The setup instructions install dependencies without version pins or hashes; this is common for a Python reporting tool but leaves package provenance and repeatability to the user.

Skill content

pip install openpyxl pandas

Recommendation

Pin dependency versions, install from trusted indexes, and prefer a reviewed requirements file or lockfile for repeatable installs.

Findings (1)

critical

suspicious.dynamic_code_execution

Location: scripts/template_engine.py:141
Finding: Dynamic code execution detected.