apple-notes-incident-runbook

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXPOSURE]: The skill accesses macOS unified logs and diagnostic reports located in ~/Library/Logs/DiagnosticReports/. These logs are used to identify the root cause of application crashes and synchronization issues. The skill explicitly instructs the agent to use redacted diagnostic locations and restricted access to manage potential exposure of sensitive data.- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection by ingesting untrusted data from external sources, including AppleScript (osascript) return values, iCloud daemon status via brctl, and macOS system logs. Evidence of this ingestion is found in the detection and diagnosis blocks of SKILL.md. The skill provides manual mitigation by advising on the classification and redaction of these diagnostics before processing. Capabilities associated with this ingestion include standard file system reads and process management tools like killall and open.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:44 AM
Security Audit — agent-trust-hub — apple-notes-incident-runbook