ingest-entra-directory-audit-ocsf

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill transforms untrusted Microsoft Entra audit logs into normalized OCSF records, creating a surface for indirect prompt injection if the agent subsequently processes this output as instructions.
  • Ingestion points: External audit data is ingested through sys.stdin or local files via the iter_raw_events function in src/ingest.py.
  • Boundary markers: The output is emitted as raw JSONL without specific delimiters or instructions to ignore embedded content.
  • Capability inventory: The script is restricted to local data transformation and does not use network requests, subprocesses, or file system writes.
  • Sanitization: Validation is limited to JSON schema integrity and a whitelist of supported activity names; field values are not sanitized for malicious natural language patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 01:17 AM
Security Audit — agent-trust-hub — ingest-entra-directory-audit-ocsf