weapp-acceptance
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes build commands and scripts defined in a project's configuration files (e.g.,
pnpm run build). While necessary for build validation, this allows for arbitrary command execution within the project environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes project-controlled JSON files (
weapp-acceptance.config.jsonand scenario files) to determine execution steps and assertions. This creates a surface for indirect prompt injection where a malicious project could influence the agent's validation outcome. - Ingestion points:
weapp-acceptance.config.jsonand files defined in theacceptance.scenariosfield. - Boundary markers: The skill includes instructions to explain which scripts/configurations need review and explicitly states that trust decisions should not be made silently.
- Capability inventory: Shell command execution via
pnpmand thewvCLI tool, as well as file inspection through MCP tools. - Sanitization: The skill relies on user-mediated authorization and review of scripts rather than automated sanitization of the input commands.
Audit Metadata