中文 EN

Scan Report

Scan New Files

✓ Low Risk — Risk Score 10/100
Last scan:18 hr ago Rescan
10 /100
scienceclaw-query
Run a scientific investigation on any topic and return findings directly to chat — without posting to Infinite
Documentation-only skill that delegates to external scienceclaw tool with no malicious patterns detected in the SKILL.md content.
Skill Namescienceclaw-query
Duration32.4s
Enginepi
✓
Safe to install
No immediate security concerns with this skill. However, actual security depends on the external $HOME/scienceclaw tool which is outside this package scope.

Findings 1 items

Severity Finding Location
Low
External tool dependency not auditable Doc Mismatch
The skill references an external tool at $HOME/scienceclaw/bin/scienceclaw-post which is not part of this package. Actual behavior depends on this external script which cannot be audited.
python3 "$SCIENCECLAW_DIR/bin/scienceclaw-post"
→ If security is critical, bundle the scienceclaw-post script within this skill package for full auditability.
 SKILL.md:16
ResourceDeclaredInferredStatusEvidence
Shell WRITE WRITE ✓ Aligned SKILL.md:16 - python3 execution
Environment READ READ ✓ Aligned SKILL.md:4 - ANTHROPIC_API_KEY
Filesystem READ READ ✓ Aligned SKILL.md:60 - memory.md reading
Network NONE UNKNOWN ✓ Aligned External scienceclaw tool may make network calls

File Tree

1 files · 3.2 KB · 77 lines
Markdown 1f · 77L
└─ 📝 SKILL.md Markdown 77L · 3.2 KB

Security Positives

✓ Uses --dry-run flag to prevent unintended posting
✓ No base64, eval, or obfuscation patterns in documentation
✓ No credential harvesting or exfiltration patterns detected
✓ Clear parameter documentation with no hidden functionality
✓ No access to sensitive paths like ~/.ssh, ~/.aws, or .env declared