resonate-basic-ephemeral-world-usage-typescript

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references the resonate-pg schema located in the resonatehq/resonate-pg GitHub repository for users choosing to run durable execution directly on Postgres without a server process.
  • [DATA_EXFILTRATION]: Instructions describe using RESONATE_TOKEN (a JWT bearer token) for server authentication and DATABASE_URL for Postgres connections. These are documented as standard environment variable patterns for the SDK's configuration.
  • [PROMPT_INJECTION]: The skill documents an attack surface for indirect prompt injection where the agent is instructed to process data from external sources.
  • Ingestion points: Untrusted data enters the agent context through resonate.promises.resolve (which accepts a data payload from webhooks or UIs) and function arguments passed to resonate.run or resonate.rpc (SKILL.md).
  • Boundary markers: None identified. There are no explicit instructions to use delimiters or warnings to ignore instructions embedded within the processed data.
  • Capability inventory: The skill has the capability to execute local and remote functions (run, rpc), write to databases (via PostgresNetwork), and interact with the Resonate Server API.
  • Sanitization: None identified. The code examples demonstrate direct interpolation of data (e.g., JSON.stringify payloads) into the application flow without sanitization or validation logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 06:42 AM
Security Audit — agent-trust-hub — resonate-basic-ephemeral-world-usage-typescript