Back to skill
Skillv1.0.0
ClawScan security
Gate Exchange Unified Account · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 6, 2026, 10:59 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's requirements and instructions are consistent with a Gate unified-account integration: it only asks for Gate API credentials, contains no installable code, and enforces confirmation before any high‑risk mutations.
- Guidance
- This skill appears coherent for Gate unified-account tasks, but it will be able to perform high‑impact operations (borrow, repay, switch modes) once you confirm them. Before enabling it: 1) provision an API key with the minimum necessary scopes and, if available, limit IP/permission scope; 2) test in read-only mode first (ask only queries) to verify behavior; 3) never paste secrets into chat (the skill forbids it); 4) when asked to confirm a mutation, carefully review the Action Draft shown by the assistant (amount, currency, risk note) before replying with confirmation; and 5) monitor your Gate account activity/logs and be prepared to revoke the API key if you see unexpected calls.
Review Dimensions
- Purpose & Capability
- okThe declared purpose (unified account equity, borrowing, leverage, mode switching) matches the requested environment variables (GATE_API_KEY, GATE_API_SECRET), declared API permissions (Unified:Read/Write), and the set of named MCP tools. No unrelated credentials, binaries, or config paths are requested.
- Instruction Scope
- okSKILL.md contains explicit, narrow runtime instructions: use the local Gate MCP session, call only the listed read/write tools, perform pre-checks, and require explicit user confirmation before any mutation. It does not ask to read unrelated files or to exfiltrate data, and it forbids pasting secrets into chat.
- Install Mechanism
- okThere is no install spec and no code files to be written or executed on the host. The skill is instruction-only, which minimizes installation risk.
- Credentials
- okOnly two env vars are required (API key and secret) which are appropriate for an exchange API integration. The primary credential is declared. The skill requests Unified:Write permission (reasonable given it supports mutations) and documents confirmation gates for high-risk actions.
- Persistence & Privilege
- okThe skill is not always-enabled, does not request system-wide persistence, and explicitly states it does not store or paste secrets. Autonomous invocation is allowed (platform default) but mutations still require explicit user confirmation.