glycoengineering

SAFE max severity MEDIUMSkillSpector Caution 47 registry K-Dense-AI last scanned 2026-10-21 1 scan

Verdict safe, but 4 findings (3 medium, 1 low), mainly data exfiltration, command injection, unauthorized tool use. Most recent scan 2026-10-21.

Findings4

Critical0

High0

Medium3

Low1

Analyzersstatic bytecode pipeline correlation behavioral trigger

Scanner version2.3.1

Second scanner: NVIDIA SkillSpector

VerdictCaution 47

Score47 / 100

Issues6

Categoriesanalysis-evasion Data Exfiltration Memory Poisoning

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

Analyzes and engineers protein glycosylation by scanning canonical N-glycosylation sequons, describing S/T-rich regions, checking curated glycan evidence, and preparing NetNGlyc, NetOGlyc and GlycoSHIELD workflows. Use for glycoprotein engineering, antibody Fc glycosylation, glycan shielding, and site-specific glycoproteomics interpretation.

Package: 5 files, 34 KB
SKILL.md: 1382 words, 11 sections
Code: Python (2)
License: Unknown
Compatibility: Local sequence helpers require Python 3.10+. Public GlyTouCan lookup requires requests and network access. DTU predictors use a browser; optional GlycoSHIELD needs a separate source installation, glycan libraries, and GROMACS for SASA analysis.

SKILL.md covers: Glycoengineering · When to use · Workflow · Local sequence analysis · Three intended changes: A1N, P2A, A3T. P->A must be explicit. · Density is a fraction of residues, not an O-glycosylation probability. · Prediction services · Structural shielding with GlycoSHIELD · Database evidence and glycan notation · Antibody engineering decisions · Experimental interpretation and verification boundary

Findings

MEDIUM Python code block sends HTTP POST request Data Exfiltrationbehavioral

Code block in references/glycan_databases.md at line 63 contains potentially dangerous Python code.

references/glycan_databases.md:63

response = requests.post(

Remediation: Review the code block for security implications.

MEDIUM Python code block executes shell commands Command Injectionbehavioral

Code block in references/glycoshield.md at line 98 contains potentially dangerous Python code.

references/glycoshield.md:98

subprocess.run(["gmx", "--version"], check=True, capture_output=True, text=True)

Remediation: Review the code block for security implications.

MEDIUM Outbound network request primitives that can transmit data externally Data Exfiltrationstatic

Pattern detected: requests.get(

scripts/glytoucan_lookup.py:34

response = requests.get(

Remediation: Ensure network operations are necessary and document allowed destinations

LOW Script contacts a destination the documentation never mentions Unauthorized Tool Usestatic

Script contacts or configures 'purl.jp' but the skill documentation never visibly mentions it. Undocumented network destinations can indicate behaviour the manifest and description do not disclose.

scripts/glytoucan_lookup.py:20

Remediation: Document the destination in SKILL.md (or the manifest description) if it is intended, or remove the undisclosed network access. Organizations can list internal hosts in llm_analysis.trusted_reference_domains.

What is in the package

File	Kind	Size
`SKILL.md`	doc	12 KB
`references/glycan_databases.md`	doc	8.2 KB
`references/glycoshield.md`	doc	6.4 KB
`scripts/glycoengineering_tools.py`	script	4.8 KB
`scripts/glytoucan_lookup.py`	script	2.6 KB

External hosts referenced

gitlab.mpcdf.mpg.de pubmed.ncbi.nlm.nih.gov glyconnect.expasy.org rdf.glytoucan.org doi.org services.healthtech.dtu.dk ts.glytoucan.org unicarb-dr.glycosmos.org api.glytoucan.org code.glytoucan.org code.google.com glytoucan.org manual.gromacs.org purl.jp support.proteinmetrics.com www.cazy.org www.genome.jp www.glytoucan.org www.matrixscience.com www.ncbi.nlm.nih.gov www.proteinmetrics.com

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.

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

Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
Bytecode analyzer.
Pipeline analyzer.
Correlation analyzer.
Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.

What we did not check

Runtime behaviour. The skill was not executed; findings come from reading its files.
Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
Content the analyzers cannot read: images, binaries, and some non-English text.
Known scanner bypass classes. See the methodology and limitations page.

Sources

Repository: https://github.com/K-Dense-AI/scientific-agent-skills
Raw scan result: skills-glycoengineering-scan.json
Badge for your README: ![skillscan.io](https://skillscan.io/badge/k-dense-ai-glycoengineering.svg)
Request a re-scan