gate-info-macroimpact
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads version metadata and updates from the official Gate GitHub repository (github.com/gate/gate-skills). These actions are part of documented maintenance workflows for the vendor's tools.
- [COMMAND_EXECUTION]: In Step 0 of SKILL.md, the agent is instructed to execute a shell-based discovery script to probe the local filesystem for the presence of the gate-info-research skill, facilitating proper tool routing.
- [DYNAMIC_EXECUTION]: The skill utilizes bash and PowerShell scripts (update-skill.sh/ps1) to manage its own lifecycle by downloading and applying updates. These scripts are executed by the agent at runtime under specific maintenance conditions.
- [INDIRECT_PROMPT_INJECTION]: The skill processes results from news feeds and economic calendars via MCP tools, creating a potential surface for adversarial instructions embedded in external content.
- Ingestion points: Data retrieved from news_feed_search_news and info_macro_get_economic_calendar in SKILL.md.
- Boundary markers: No explicit delimiters or boundary markers are defined to isolate untrusted tool output.
- Capability inventory: The skill possesses the capability to execute shell scripts and perform filesystem discovery.
- Sanitization: External content is interpolated into the analysis report without explicit sanitization or filtering logic.
Audit Metadata