gate-news-eventexplain
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches updates and configuration from the official GitHub repository for the vendor 'gate'. This includes downloading skill files via git, ZIP, or tarball formats as part of the skill's self-maintenance workflow.\n- [COMMAND_EXECUTION]: Executes local maintenance scripts to perform version checks and apply updates. It also performs a shell probe to detect the presence of the
gate-clitool and relevant skill directories for routing purposes.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user queries and external data (news headlines, event details, and on-chain metrics) to generate analysis reports.\n - Ingestion points: Ingests coin symbols, time hints, and results from news_events_get_latest_events, news_feed_search_news, and info_onchain_get_token_onchain.\n
- Boundary markers: Utilizes a structured Markdown report template to separate analysis segments.\n
- Capability inventory: The skill possesses file-writing capabilities via update scripts and network-reading capabilities via MCP tools.\n
- Sanitization: Relies on instructional reasoning to identify trigger events and verify causality, with no explicit programmatic sanitization of external data mentioned.
Audit Metadata