skills/mukul975/anthropic-cybersecurity-skills/performing-active-directory-compromise-investigation/Gen Agent Trust Hub
performing-active-directory-compromise-investigation
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it ingests untrusted Windows event log data to generate investigation reports.
- Ingestion points: The scripts
scripts/agent.py(JSON) andscripts/process.py(XML) read external log files provided via CLI arguments. - Boundary markers: There are no delimiters or instructions provided to the agent to treat the log content as data rather than potential instructions.
- Capability inventory: The skill utilizes file-system write access across both
agent.pyandprocess.pyto create JSON report files. No network operations or administrative shell capabilities were identified. - Sanitization: The log data is parsed and aggregated without sanitization or escaping of the field values before being written to the output reports.
- [SAFE]: The
scripts/process.pyfile uses the standardxml.etree.ElementTreelibrary to parse XML event logs. This library does not protect against XML External Entity (XXE) attacks, which could be exploited by a malicious log file to perform unauthorized local file reads.
Audit Metadata