instantly-hello-world
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references official Instantly packages and documentation, including the '@instantlyai/sdk' and '@instantlyai/cli' npm packages, and various developer guides from the developer.instantly.ai domain. These are recognized as legitimate resources from a well-known service provider.\n- [INDIRECT_PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks due to its capability to ingest data from an external API and modify local files.\n
- Ingestion points: Data is ingested from the Instantly API via the 'WebFetch' tool at 'https://api.instantly.ai/api/v2'.\n
- Boundary markers: The skill includes explicit instructions to 'Capture latency, status, route template, and redacted item count' and 'sensitive fields redacted', establishing clear boundaries for data handling.\n
- Capability inventory: The skill is granted 'Write' and 'Edit' tools, which allow for local file modification based on the processed external data.\n
- Sanitization: The instructions mandate failing closed on schema or tenant mismatches and specifically forbid printing, persisting, or committing credentials, while requiring redaction of items in the output handoff.
Audit Metadata