functions
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides official documentation for the Zavu Cloud platform. All external dependencies (@zavudev/functions, @zavudev/sdk) and CLI tools (zavudev) are trusted vendor resources provided by the author.
- [INDIRECT_PROMPT_INJECTION]: The skill outlines an architecture where agents ingest and process external data via tool handlers and event triggers.
- Ingestion points: Data enters the system through tool arguments in
defineTooland inbound events indefineFunction. - Boundary markers: While the platform validates tool arguments against a schema, the provided examples do not explicitly show manual delimiter use for prompt inputs.
- Capability inventory: The defined handlers have the capability to perform network operations (
fetch), interact with persistent storage (ctx.memory), and execute logic in a serverless environment. - Sanitization: Tool arguments are automatically validated against a developer-defined JSON schema before execution.
Audit Metadata