Back to skill
Skillv1.0.0
ClawScan security
Core PA Admin and Exec Support · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 4:50 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This instruction-only skill is coherently limited to turning user-provided admin context into plans, drafts, and scheduling proposals without taking actions.
- Guidance
- This appears safe to install as an instruction-only drafting and planning aid. Review generated drafts and schedules before using them, and only paste messages, calendar details, or meeting notes that you are comfortable sharing with the assistant.
Review Dimensions
- Purpose & Capability
- okThe stated purpose matches the artifacts: it ingests pasted messages, calendars, tasks, and notes to produce plans, drafts, agendas, briefs, and JSON output.
- Instruction Scope
- okThe instructions explicitly prohibit final actions such as sending emails or booking meetings and require STOP-and-ASK when key information is missing.
- Install Mechanism
- okThere is no install spec, no required binaries, no environment variables, and no code files, so there is no artifact-backed execution or dependency risk.
- Credentials
- okThe skill requests no credentials, account access, network tools, filesystem access, or OS-specific permissions; its authority is proportionate to drafting and planning.
- Persistence & Privilege
- okThe artifacts show no persistence, background execution, credential use, memory storage, or autonomous activity beyond producing user-facing drafts and proposals.