asta-documents

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Installs the asta CLI tool and its associated plugins from the official Allen Institute for AI (allenai) GitHub repository.
  • [EXTERNAL_DOWNLOADS]: Downloads remote document index files and content from external URLs and cloud storage providers (S3, GCS) as part of its document management workflow.
  • [COMMAND_EXECUTION]: Uses python3, curl, and sed within provided bash instructions to parse URL components and download remote index files.
  • [DATA_EXFILTRATION]: Provides functionality to access local files via the file:// protocol for indexing and retrieval purposes.
  • [PROMPT_INJECTION]: The skill ingests metadata from external index files and documents (Ingestion points: asta:// URLs and .asta/ directory files in SKILL.md) which could serve as an indirect prompt injection surface. Boundary markers: absent. Capability inventory: Bash, Read, Write in SKILL.md. Sanitization: absent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 04:01 AM
Security Audit — agent-trust-hub — asta-documents