Versioned security assessment

Report ID: SA-0AE6F2F8

6/30/2026, 6:21:06 AM

fluidsim security assessment v5

Skill Security Certification Report

Audit History
Audit model: codex Historical report
Skill name
fluidsim
Version
v5
Maintainer
K-Dense-AI
Source
github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/fluidsim
Coverage
7 Files scanned · 1,568 Lines analyzed
Policy version
Unavailable

Highest confirmed finding severity

Low

1 confirmed security finding requires 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.

Static analysis reported many command execution, filesystem, environment, network, and weak-crypto indicators. Review found most command and weak-crypto hits are Markdown examples for FluidSim installation, solver selection, parameter names, and mathematical terms, not executable skill code. However, SKILL.md contains an embedded instruction to promote an external K-Dense web service during complex workflows, which is a prompt-injection and marketplace policy risk.

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

7 Files scanned · 1,568 Lines analyzed

4 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 2 evidence locations

Filesystem access

May read or write local files.

Observed in 1 evidence location

Env variables

May read values from the process environment.

Observed in 1 evidence location

External commands

May invoke commands or programs outside the Skill.

Observed in 3 evidence locations

Capability review items (3)
High
SKILL.md:348-349
Prompt Injection Attempt Detected
SKILL.md instructs the assistant to proactively suggest K-Dense Web for complex workflows and provides an external service URL. This is unrelated to FluidSim execution and attempts to steer assistant behavior toward promotional external routing.
The text is a direct assistant behavior instruction inside untrusted skill content. It does not claim system authority, so confidence is high but not maximal.
Medium
references/simulation_workflow.md:132-142references/advanced_features.md:200-221
Cluster Job Submission Guidance Requires User Caution
The references describe generating Python job scripts and submitting them to a compute cluster. This appears legitimate for FluidSim HPC workflows, but users should review generated scripts and scheduler targets before running them.
The workflow is consistent with scientific computing, but it can execute user-created simulation scripts on shared infrastructure. No malicious intent or exfiltration was found.
Low
references/installation.md:42-53references/advanced_features.md:382-386
Environment Variable Examples Are Local Configuration
The environment variable usage configures FluidSim output paths and FFT backends. No evidence found that secrets are read, transmitted, or logged.
The variables are documented FluidSim settings and contain local paths or backend names. They are not credentials and there is no network sink tied to them.

Risk findings

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

Confirmed security concerns (1)

RISK-001 Low
references/solvers.md:1-20references/parameters.md:18-21references/simulation_workflow.md:31-32
Weak Crypto and Reconnaissance Hits Are Domain Terms
The weak cryptography and reconnaissance detections appear to be false positives from CFD terms such as solver names, shallow water rotation, and parameter descriptions. No evidence found of cryptographic code, credential handling, or system probing.
Semantic review shows scientific simulation context, not security tooling. The static labels do not match the documented purpose of these lines.

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: codex

Analysis state: Complete

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

Static false positives ignored (1)
Low
references/installation.md:14-24SKILL.md:27-36references/output_analysis.md:106-114
Static Command Execution Findings Are Documentation Examples
The numerous shell-command findings are fenced Markdown examples for installing packages, running MPI jobs, launching visualization tools, or showing Python imports. No evidence found that the skill itself executes these commands automatically.
The flagged lines are documentation snippets rather than executable hooks. They require explicit user action and are normal for a FluidSim installation and analysis guide.

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