abridge-hello-world
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from the local repository (configuration, adapters, tests) and external documentation via
WebFetch. - Ingestion points: File
SKILL.md(Steps 2 and 6) and external links provided inreferences/official-docs.md. - Boundary markers: Present. The skill explicitly instructs the agent to record outcomes without clinical text and to only return exact operator steps or approval-gated handoffs for live actions rather than executing them autonomously.
- Capability inventory: The skill utilizes
Read,Glob,Grep,WebFetch,Write, andEdittools. - Sanitization: Present. The instructions mandate confirming environment scope and organizational approval states before using write or edit tools, alongside strict prohibitions against including clinical data in logs.
- [EXTERNAL_DOWNLOADS]: Fetches product best practices, recording basics, and integration guidance from Abridge's official support documentation at
support.abridge.com.
Audit Metadata