gate-exchange-newcoin

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the gate-cli binary and associated checksums from the vendor's official GitHub repository.
  • Evidence: setup.sh and SKILL.md frontmatter reference https://github.com/gate/gate-cli/releases/download/v0.6.2/ for platform-specific binaries.
  • [COMMAND_EXECUTION]: Executes the gate-cli tool to interact with exchange APIs for research and trading purposes.
  • Evidence: SKILL.md documents 12 specific CLI commands under the info, news, and cex namespaces used to fetch market data and manage orders.
  • [SAFE]: Implements mandatory safety guardrails for all sensitive operations.
  • Evidence: SKILL.md and references/gate-runtime-rules.md enforce a human-in-the-loop confirmation gate (Action Draft + Y/N) before any order is placed via gate-cli cex spot order.
  • [PROMPT_INJECTION]: Analyzed the attack surface for indirect prompt injection from external exchange data and news feeds.
  • Ingestion points: Data is ingested through gate-cli news feed get-exchange-announcements and search-news as defined in SKILL.md.
  • Boundary markers: Present. Mandatory Action Drafts with risk disclosures and explicit user confirmation are required for all execution steps.
  • Capability inventory: Order placement capabilities are restricted to gate-cli cex spot order and gate-cli cex alpha order place as defined in SKILL.md.
  • Sanitization: Present. The skill uses structured JSON output (--format json) for data parsing to prevent instruction confusion.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 11:29 AM