zustellbarkeit

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted email content from EML files, text, or HTML, creating a potential surface for indirect prompt injection attacks. 1. Ingestion points: Email content is ingested through file reading and CLI arguments as defined in SKILL.md and scripts/zustellbarkeit.py. 2. Boundary markers: SKILL.md includes binding protection rules (Rules 6 and 7) instructing the agent to treat processed data as non-instructional material and to avoid following embedded links. 3. Capability inventory: The skill is limited to reading local files and executing its own script; it has no network, write, or privilege escalation capabilities. 4. Sanitization: The script extracts technical metadata, and the instructions act as a cognitive filter for the agent against malicious directives in the data.
  • [SAFE]: The skill relies on a local script using only the Python standard library. It does not perform network requests, download external code, or require elevated privileges. All analysis occurs within the local environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 12:16 PM
Security Audit — agent-trust-hub — zustellbarkeit