中文 EN

Scan Report

Scan New Files

✓ Trusted — Risk Score 5/100
Last scan:19 hr ago Rescan
5 /100
daily-literature
Automated daily literature search system for academic researchers
This is a legitimate academic literature search skill with no malicious behavior detected. The subprocess call to another skill is a documented inter-skill communication pattern for literature retrieval.
Skill Namedaily-literature
Duration33.6s
Enginepi
✓
Safe to install
No action required. The skill is safe to use.

Findings 1 items

Severity Finding Location
Low
Minor documentation gap for subprocess usage Doc Mismatch
The SKILL.md header does not explicitly declare shell:WRITE capability, though the main documentation body describes the search_lit_review function and inter-skill communication. This is a minor documentation style issue rather than malicious concealment.
env_vars section missing shell capability declaration
→ Add allowed-tools declaration to SKILL.md header if subprocess is intentional
 SKILL.md:1
ResourceDeclaredInferredStatusEvidence
Filesystem READ READ ✓ Aligned File read/write limited to workspace directory
Network READ READ ✓ Aligned API calls to PubMed, OpenAlex, Semantic Scholar, Unpaywall
Shell NONE WRITE ✓ Aligned daily_literature_search.py:210 - subprocess.run() for inter-skill communication
Environment READ READ ✓ Aligned Uses os.getenv() for documented API keys
Skill Invoke NONE READ ✓ Aligned Invokes literature-review skill for search functionality
8 findings
🔗
Medium External URL 外部 URL
https://api.unpaywall.org/v2/
scripts/daily_literature_search.py:431
🔗
Medium External URL 外部 URL
https://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi
scripts/daily_literature_search.py:498
🔗
Medium External URL 外部 URL
https://www.ncbi.nlm.nih.gov/pmc/articles/
scripts/daily_literature_search.py:518
🔗
Medium External URL 外部 URL
https://doi.org/
scripts/daily_literature_search.py:594
🔗
Medium External URL 外部 URL
http://doi.org/
scripts/utils.py:191
🔗
Medium External URL 外部 URL
https://doi.org/10.1038/s41375-026-02867-9
tests/test_classifier.py:61
🔗
Medium External URL 外部 URL
https://doi.org/10.1007/s11899-026-00772-7
tests/test_classifier.py:76
📧
Info Email 邮箱地址
[email protected]
README.md:131

File Tree

10 files · 76.8 KB · 2761 lines
Python 5f · 1673L Markdown 2f · 622L Shell 1f · 274L YAML 1f · 178L Text 1f · 14L
├─ ▾ 📁 scripts
│ ├─ 🐍 analyze_uploaded.py Python 277L · 8.1 KB
│ ├─ 🐍 classifier.py Python 167L · 4.7 KB
│ ├─ 🐍 daily_literature_search.py Python 821L · 25.8 KB
│ └─ 🐍 utils.py Python 258L · 7.4 KB
├─ ▾ 📁 tests
│ └─ 🐍 test_classifier.py Python 150L · 5.2 KB
├─ 📋 config.example.yaml YAML 178L · 3.8 KB
├─ 🔧 install.sh Shell 274L · 6.9 KB
├─ 📝 README.md Markdown 373L · 7.9 KB
├─ 📄 requirements.txt Text 14L · 271 B
└─ 📝 SKILL.md Markdown 249L · 6.7 KB

Dependencies 2 items

PackageVersionSourceKnown VulnsNotes
requests >=2.28.0 pip No Version pinned with minimum
pyyaml >=6.0 pip No Version pinned with minimum

Security Positives

✓ Dependencies are version-pinned (requests>=2.28.0, pyyaml>=6.0)
✓ All network requests go to legitimate academic APIs (PubMed, OpenAlex, Semantic Scholar, Unpaywall)
✓ No credential exfiltration - API keys are used locally only for authentication
✓ File operations are scoped to workspace directory only
✓ No access to sensitive paths like ~/.ssh, ~/.aws, or .env files
✓ No obfuscation techniques (base64, eval, atob) detected
✓ No reverse shell, C2, or data theft patterns found
✓ No suspicious cron/scheduled task modifications
✓ Uses subprocess to invoke another legitimate skill (literature-review) for search functionality