Back to skill
Skillv1.0.0
ClawScan security
PubMed · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 5:32 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This instruction-only PubMed skill provides search and appraisal guidance without requesting credentials, installing code, or asking for unusual system access.
- Guidance
- This appears safe to install as an instruction-only research aid. Users should still treat biomedical literature summaries as informational support and rely on qualified medical professionals for clinical decisions.
Review Dimensions
- Purpose & Capability
- okThe content is coherent with the stated purpose: it gives biomedical literature search, filtering, study-quality, and critical-appraisal guidance.
- Instruction Scope
- okInstructions are limited to research workflow advice and do not direct the agent to override user intent, run commands, mutate accounts, or handle sensitive local data.
- Install Mechanism
- okThe artifacts state there is no install spec and no code files; the skill is instruction-only.
- Credentials
- okNo binaries, environment variables, credentials, config paths, or capability tags are requested.
- Persistence & Privilege
- okNo persistence, background operation, privileged access, credential use, or stored memory behavior is evidenced in the artifacts.