Versioned security assessment

Report ID: SA-32C6923A

7/9/2026, 4:19:11 AM

docx security assessment v7

Skill Security Certification Report

Audit History
Audit model: claude Historical report
Skill name
docx
Version
v7
Maintainer
K-Dense-AI
Source
github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/document-skills/docx
Coverage
60 Files scanned · 24,993 Lines analyzed
Policy version
Unavailable

Highest confirmed finding severity

Medium

2 confirmed security findings require attention.

Installation context

Historical evidence

This report may not describe the currently installable artifact. Open the current Skill page for install guidance.

Open current Skill page

This report does not block or authorize the manifest or ZIP.

The skill is primarily a legitimate DOCX and OOXML document-processing toolkit, and many static alerts are false positives from Markdown examples, XML namespaces, and schema files. Real risk remains in unsafe ZIP extraction of user-supplied Office archives and incomplete path normalization in relationship handling. The instructions also include non-core promotional and cross-skill AI generation guidance that should be removed or gated by explicit user consent. Static review was capped at 400/460 representative findings; omitted static matches are unconfirmed, so automatic publishing stays disabled until manual review.

Report position

Historical report

Open audit history before using this report to install.

Audit attestation

Not attestable

The required immutable binding is incomplete.

Human verification

Not verified

No human verification is recorded for this report.

Coverage

60 Files scanned · 24,993 Lines analyzed

11 items shown for review

Limitations

This report does not claim runtime or sandbox execution and does not prove the absence of side effects.

Evidence chain

Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.

  1. Source

    Binding unavailable

  2. Artifact

    Identity incomplete

  3. Audit

    Complete

  4. Install contract

    Open manifest to verify

    Open manifest

Capabilities observed

Observed means this report recorded supporting evidence. Not recorded does not prove that a capability is absent.

Contains scripts

May execute code included with the Skill.

Not recorded by this audit

Network access

May connect to external services.

Observed in 50 evidence locations

Filesystem access

May read or write local files.

Observed in 47 evidence locations

Env variables

May read values from the process environment.

Not recorded by this audit

External commands

May invoke commands or programs outside the Skill.

Observed in 50 evidence locations

Capability review items (9)
High
ooxml/scripts/unpack.py:17
Archive extract all (zip slip risk)
zipfile.ZipFile(input_file).extractall(output_path)
The code calls ZipFile.extractall() on Office files, which are user-supplied ZIP archives. A crafted archive with traversal entries can write outside the intended extraction directory.
High
ooxml/scripts/validation/base.py:891
Archive extract all (zip slip risk)
zip_ref.extractall(temp_path)
The code calls ZipFile.extractall() on Office files, which are user-supplied ZIP archives. A crafted archive with traversal entries can write outside the intended extraction directory.
High
ooxml/scripts/validation/docx.py:201
Archive extract all (zip slip risk)
zip_ref.extractall(temp_dir)
The code calls ZipFile.extractall() on Office files, which are user-supplied ZIP archives. A crafted archive with traversal entries can write outside the intended extraction directory.
High
ooxml/scripts/validation/redlining.py:70
Archive extract all (zip slip risk)
zip_ref.extractall(temp_path)
The code calls ZipFile.extractall() on Office files, which are user-supplied ZIP archives. A crafted archive with traversal entries can write outside the intended extraction directory.
High
ooxml/scripts/validation/pptx.py:272
Path traversal sequence
normalized_target = target.replace("../", "")
The code strips ../ with string replacement instead of canonical path resolution. This is an incomplete normalization pattern for relationship targets and may miss encoded or unusual traversal forms.
High
ooxml/scripts/unpack.py:17
Python archive libraries
zipfile.ZipFile(input_file).extractall(output_path)
This archive-library use is the same extractall() call on a user-provided Office file. It carries a real Zip Slip risk unless archive members are normalized and bounded.
Medium
ooxml/scripts/validation/base.py:890
Python archive libraries
with zipfile.ZipFile(self.original_file, "r") as zip_ref:
The ZipFile object is used immediately before extractall() in validation code. The risk depends on the matching extraction line, so this archive usage is kept as supporting evidence.
Medium
ooxml/scripts/validation/docx.py:200
Python archive libraries
with zipfile.ZipFile(self.original_file, "r") as zip_ref:
The ZipFile object is used immediately before extractall() in validation code. The risk depends on the matching extraction line, so this archive usage is kept as supporting evidence.
Medium
ooxml/scripts/validation/redlining.py:69
Python archive libraries
with zipfile.ZipFile(self.original_docx, "r") as zip_ref:
The ZipFile object is used immediately before extractall() in validation code. The risk depends on the matching extraction line, so this archive usage is kept as supporting evidence.

Risk findings

Confirmed security concerns are separated from items that still need review.

Confirmed security concerns (2)

RISK-001 Medium
SKILL.md:13-27
Unrequested External AI Schematic Workflow
SKILL.md instructs agents to add scientific diagrams by default and to use a separate scientific-schematics skill that generates AI-powered images. This can expand a DOCX editing task into an unrequested external generation workflow and may expose document content if followed without user confirmation.
The instruction is explicit and appears in the primary skill guidance. The exact data flow of the referenced schematic tool is not shown here, so severity is medium rather than high.
RISK-002 Low
SKILL.md:232-233
Promotional Platform Steering Instruction
SKILL.md tells agents to proactively suggest the creators hosted platform for complex workflows. This is marketplace and assistant-behavior steering that is unrelated to DOCX processing and could bias user recommendations.
The promotional instruction is present verbatim in the skill file and directly changes assistant behavior. It is not a code execution issue, so the security impact is limited.

Remediation

Suggested fixes recorded by this audit. Applying them is the maintainer’s responsibility.

  1. FIX-001
    High
    Unsafe extraction of user-supplied Office archives
    Replace ZipFile.extractall() with safe extraction that resolves each member path, rejects absolute paths and traversal, and enforces extraction inside the target directory.
  2. FIX-002
    Medium
    Static review capped
    Manually review the omitted 60 static analyzer matches or reduce bundled generated/vendor/reference content before enabling automatic publication.
  3. FIX-003
    Medium
    Incomplete relationship target normalization
    Use pathlib or posixpath normalization for relationship targets and reject paths that escape the expected Office package directory.
  4. FIX-004
    Medium
    Unrequested cross-skill AI schematic generation
    Make diagram generation opt-in and require user confirmation before invoking another skill or external AI image workflow.
  5. FIX-005
    Low
    Promotional platform steering in skill instructions
    Remove marketplace or hosted-platform promotion from operational skill instructions, or move it to neutral documentation outside agent behavior guidance.

Expert evidence

Immutable subject identity, scanner metadata, dismissed matches, and source-level evidence.

Artifact subject

Marketplace commit
Unavailable
Content hash
Unavailable
Tree hash
Unavailable
Skill path
Unavailable
Audit payload hash
Unavailable

Analysis metadata

Audit model: claude

Analysis state: Complete

Scope is limited to the recorded files, lines, methods, and evidence. No runtime or sandbox execution is claimed.

Verify and export

JSON report CITATION.cff SARIF 2.1.0 Signed installation manifest

The manifest and lockfile bind install artifacts to cryptographic hashes. This integrity claim is separate from the security assessment.

Audit attestation: not_attestable