Back to skill
Skillv0.1.0
ClawScan security
Architecture Designer · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 4:55 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This is an instruction-only architecture guidance skill with no code, install steps, credentials, persistence, or external data access shown in the artifacts.
- Guidance
- This skill appears safe to install as a documentation and architecture-planning helper. Treat its outputs as design recommendations and continue to review architecture decisions with appropriate stakeholders, especially for security, cost, and operational trade-offs.
Review Dimensions
- Purpose & Capability
- okThe stated purpose is architecture design and review, and the included content is limited to design workflows, ADR templates, architecture patterns, database selection, NFRs, and system design guidance.
- Instruction Scope
- okRuntime instructions are advisory and scoped to architecture analysis and documentation; they do not direct the agent to override user intent, run tools, access private data, or take external actions.
- Install Mechanism
- okThere is no install specification and no code files; the supplied files are markdown references only.
- Credentials
- okThe skill declares no required binaries, environment variables, credentials, config paths, or OS-specific access, which is proportionate for an instruction-only design assistant.
- Persistence & Privilege
- okNo persistence mechanism, background behavior, privileged access, memory storage, or credential use is described in the artifacts.