hebrew-tech-lecture-summary
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from various sources such as user-pasted text, transcription files, and external links to generate summaries. Malicious instructions embedded in this source material could potentially influence the agent's behavior. Ingestion points include user input and files like
output/transcription_plain.txt. There are no explicit boundary markers or delimiters specified to distinguish source content from agent instructions. No sanitization or validation of the input text is performed before processing. - [COMMAND_EXECUTION]: The skill invokes a local Python script to perform document conversion tasks. Evidence:
SKILL.mdinstructs the agent to runpython "<skill_dir>/scripts/md_to_word.py" "<path_to_summary.md>". - [EXTERNAL_DOWNLOADS]: The skill references an external dependency on the
python-docxpackage from the official Python Package Index (PyPI). Evidence:requirements.txtandSKILL.mdmentionpython-docx, which is a well-known and standard library for Word document creation.
Audit Metadata