primitive-functions

Pass

Audited by Gen Agent Trust Hub on Oct 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the vendor's CLI tool globally using npm install -g @primitivedotdev/cli. This is a standard installation procedure for the service provided by the vendor.
  • [COMMAND_EXECUTION]: The skill makes extensive use of the primitive CLI to manage function lifecycles, including initialization, deployment, secret configuration, and log retrieval. It also uses npm for building local projects.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from inbound emails, which presents a surface for indirect prompt injection if the email content is used to drive automated logic or LLM prompts.
  • Ingestion points: Untrusted data enters the system via the email.received event processed by the deployed JavaScript handlers.
  • Boundary markers: The instructions explicitly recommend using the isLoop check and advise treating the email body as untrusted input.
  • Capability inventory: Deployed functions have the capability to reply to emails, send new messages, and call external APIs using provided secrets.
  • Sanitization: The skill recommends using isTrustedSender from the vendor's SDK to verify the origin of mail before performing privileged actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 17, 2026, 10:44 PM