The Agent Skills Directory

[SAFE]: No security issues were detected. The skill consists of technical documentation and reference material for the Truto platform's JSONata extensions. All identified patterns, such as encoding/decoding and AI API integration, are standard for the intended use case of data transformation and integration mapping.
[INDIRECT_PROMPT_INJECTION]: The skill documents the processing of untrusted external data (integration responses, webhooks) using JSONata, which represents a potential attack surface. However, this is inherent to the primary function of the platform and the skill provides appropriate guidance for safe usage.
Ingestion points: references/usage-in-truto.md identifies integration responses (response.body), inbound webhook payloads, and account context as primary data sources.
Boundary markers: JSONata expressions serve as a structural transformation layer between external inputs and the agent's context.
Capability inventory: Functions include hashing ($digest, $sign), encoding ($base64encode), and AI embedding generation ($generateEmbeddingsCohere).
Sanitization: The documentation promotes the use of coercion and cleanup functions like $toNumber, $string, and $removeEmpty to handle data safely.

truto-jsonata