modal

ISSUES FOUND max severity HIGHSkillSpector Do not install 94 registry K-Dense-AI last scanned 1 scan

8 findings (3 high, 5 medium), mainly command injection, data exfiltration. Most recent scan 2026-10-21.

Findings8
Critical0
High3
Medium5
Low0
Analyzersstatic bytecode pipeline correlation behavioral trigger
Scanner version2.3.1

Second scanner: NVIDIA SkillSpector

VerdictDo not install 94
Score94 / 100
Issues10
CategoriesPrivilege Escalation analysis-evasion Excessive Agency
Version2.12.0 (static)

Independent static run, no language model, over the same files as the Cisco scan above. How the two scanners compare across the whole corpus: Two scanners, one corpus.

What this skill does

Modal is a serverless cloud platform for running Python on demand, including on-demand GPUs. Use when deploying or serving AI/ML models, running GPU-accelerated workloads (training, fine-tuning, inference), serving web endpoints, scheduling batch jobs, or scaling Python code to cloud containers with the Modal SDK.

Package
13 files, 96 KB
SKILL.md
2073 words, 12+ sections
Code
None. Instructions only.
License
Apache-2.0
Compatibility
Requires Python 3.10-3.14 and modal 1.6.0. Cloud execution needs a Modal account, authentication and network access; GPU use needs a payment method. Workload dependencies belong in their container Images.

SKILL.md covers: Modal · Overview · When to Use This Skill · Installation and Authentication · Install · Authenticate · Core Concepts · App and Functions · Container Images · GPU Compute · Multiple GPUs · GPU fallback chain

Findings

HIGH Python code block uses eval/exec Command Injectionbehavioral

Code block in SKILL.md at line 352 contains potentially dangerous Python code.

SKILL.md:352

proc = sb.exec("python", "/tmp/job.py")
Remediation: Review the code block for security implications.
HIGH Python code block uses eval/exec Command Injectionbehavioral

Code block in references/api_reference.md at line 137 contains potentially dangerous Python code.

references/api_reference.md:137

proc = sb.exec("python", "-c", "print(2 ** 10)")
Remediation: Review the code block for security implications.
HIGH Python code block uses eval/exec Command Injectionbehavioral

Code block in references/functions.md at line 101 contains potentially dangerous Python code.

references/functions.md:101

self.model.eval()  # PyTorch inference mode — not Python's built-in eval()
Remediation: Review the code block for security implications.
MEDIUM Python code block executes shell commands Command Injectionbehavioral

Code block in references/gpu.md at line 161 contains potentially dangerous Python code.

references/gpu.md:161

subprocess.run([
Remediation: Review the code block for security implications.
MEDIUM Python code block executes shell commands Command Injectionbehavioral

Code block in references/gpu.md at line 181 contains potentially dangerous Python code.

references/gpu.md:181

subprocess.run(["python", "train_script.py"], check=True)
Remediation: Review the code block for security implications.
MEDIUM Python code block executes shell commands Command Injectionbehavioral

Code block in references/gpu.md at line 192 contains potentially dangerous Python code.

references/gpu.md:192

subprocess.run([
Remediation: Review the code block for security implications.
MEDIUM Python code block sends HTTP POST request Data Exfiltrationbehavioral

Code block in references/scheduled-jobs.md at line 150 contains potentially dangerous Python code.

references/scheduled-jobs.md:150

response = requests.post(os.environ["SLACK_URL"], json={"text": f"Alert: {status}"}, timeout=30)
Remediation: Review the code block for security implications.
MEDIUM Python code block executes shell commands Command Injectionbehavioral

Code block in references/web-endpoints.md at line 168 contains potentially dangerous Python code.

references/web-endpoints.md:168

subprocess.Popen([
Remediation: Review the code block for security implications.

What is in the package

FileKindSize
SKILL.mddoc18 KB
references/api_reference.mddoc8.5 KB
references/examples.mddoc10 KB
references/functions.mddoc8.3 KB
references/getting-started.mddoc4.7 KB
references/gpu.mddoc6.5 KB
references/images.mddoc7.3 KB
references/resources.mddoc3.2 KB
references/scaling.mddoc6.5 KB
references/scheduled-jobs.mddoc4.3 KB
references/secrets.mddoc3.9 KB
references/volumes.mddoc6.8 KB
references/web-endpoints.mddoc8.2 KB

External hosts referenced

modal.com arxiv.org github.com docs.pytorch.org docs.vllm.ai doi.org export.arxiv.org huggingface.co sbert.net

Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.

Environment variables referenced

GITHUB_TOKEN HF_TOKEN MODAL_TOKEN_ID MODAL_TOKEN_SECRET AWS_ACCESS_KEY_ID AWS_SECRET_ACCESS_KEY OPENAI_API_KEY SLACK_URL WANDB_API_KEY

Package contents recorded 2026-10-22 at commit 593377bb0487. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.

What we checked

What we did not check

Sources

Related skills