scientific-schematics
5 findings (3 critical, 2 medium), mainly data exfiltration, unauthorized tool use. Most recent scan 2026-10-21.
Second scanner: NVIDIA SkillSpector
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
Generates scientific diagram drafts using Nano Banana 2 AI with smart iterative refinement. Uses Gemini 3.7 Flash for quality review. Refines when the review requests improvement, with at most two generations. Specialized in neural network architectures, system diagrams, flowcharts, biological pathways, and complex scientific visualizations.
- Package
- 6 files, 101 KB
- SKILL.md
- 2774 words, 12+ sections
- Code
- Python (2), Shell (1)
- Allowed tools
ReadWriteEditBash- License
- MIT license
- Compatibility
- Requires Python 3.10+ with requests, network access, and an OpenRouter API key.
SKILL.md covers: Scientific Schematics and Diagrams · Overview · Quick Start: Generate Any Diagram · Generate for journal paper (highest quality threshold: 8.5/10) · Generate for presentation (lower threshold: 6.5/10 - faster) · Generate for poster (moderate threshold: 7.0/10) · Custom max iterations (max 2) · Configuration · Current API contract · AI Generation Best Practices · When to Use This Skill · How to Use This Skill
Findings
CRITICAL Cross-file env var exfiltration: 2 files Data Exfiltrationbehavioral
Environment variable access with network calls in scripts/generate_schematic_ai.py, scripts/generate_schematic.py
CRITICAL Cross-file exfiltration chain: 2 files Data Exfiltrationbehavioral
Multi-file exfiltration chain detected: scripts/generate_schematic_ai.py, scripts/generate_schematic.py collect data → scripts/generate_schematic_ai.py → scripts/generate_schematic_ai.py transmit to network
CRITICAL Environment variable access with network calls detected Data Exfiltrationbehavioral
Script accesses environment variables and makes network calls in /home/runner/work/skill-scanner-test/skill-scanner-test/skills/K-Dense-AI-scientific-agent-skills/skills/scientific-schematics/scripts/generate_schematic_ai.py
/home/runner/work/skill-scanner-test/skill-scanner-test/skills/K-Dense-AI-scientific-agent-skills/skills/scientific-schematics/scripts/generate_schematic_ai.py
MEDIUM Code makes network requests Unauthorized Tool Usestatic
Skill code makes network requests. While not controlled by allowed-tools, network access should be documented and justified in the skill description.
SKILL.md
MEDIUM Outbound network request primitives that can transmit data externally Data Exfiltrationstatic
Pattern detected: requests.post(
scripts/generate_schematic_ai.py:263
response = requests.post(What is in the package
| File | Kind | Size |
|---|---|---|
SKILL.md | doc | 20 KB |
references/best_practices.md | doc | 20 KB |
references/iterative_refinement.md | doc | 15 KB |
scripts/example_usage.sh | script | 3.5 KB |
scripts/generate_schematic.py | script | 7.6 KB |
scripts/generate_schematic_ai.py | script | 35 KB |
External hosts referenced
openrouter.ai www.consort-spirit.org www.nature.com www.science.org arxiv.org jfly.uni-koeln.de research-figure-guide.nature.com www.prisma-statement.org colorable.jxnblk.com colorbrewer2.org doi.org export.arxiv.org inkscape.org journals.ieeeauthorcenter.ieee.org journals.plos.org reactome.org schemdraw.readthedocs.io webaim.org www.cell.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.
Environment variables referenced
OPENROUTER_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
- 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-scientific-schematics-scan.json
- Badge for your README:
- Request a re-scan