tinybird-typescript-sdk-guidelines
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines an environment that ingests untrusted data from external sources and processes it using SQL templates, which is a standard surface for injection attacks.
- Ingestion points: The skill defines connections to Kafka, Amazon S3, and Google Cloud Storage (GCS) as described in
rules/connections.mdandrules/copy-sink-pipes.md. - Boundary markers: SQL templating syntax
{{Type(param)}}is used inrules/defining-endpoints.mdto interpolate parameters, providing type-level validation but not complete protection against malicious logic in data. - Capability inventory: The skill enables the agent to execute workspace-wide CLI commands (
tinybird deploy,tinybird build), perform network operations to Tinybird Cloud, and manage sensitive tokens. - Sanitization: The SDK provides parameter modifiers (e.g.,
p.int32(),p.dateTime()) that enforce schema validation on user-supplied inputs.
Audit Metadata