paper-micro-unit-generator
Warn
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The script
scripts/generate_all_offline.pyuses theexec()function to process the contents ofstep3_filled_placeholder.py. This allows for the execution of arbitrary Python code contained within that file during the skill's operation. - Evidence:
exec(p.read_text(encoding="utf-8"), ns)inscripts/generate_all_offline.py. - [PROMPT_INJECTION]: The skill processes multiple external data sources to assemble a final document, creating a surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points:
paper_output/tasks.json,step3_filled_placeholder.py,A_题意对齐.md, and other workspace documents. - Boundary markers: Absent. The skill does not use delimiters or instructions to distinguish data from instructions in the ingested files.
- Capability inventory: File system writing in
scripts/generate_all_offline.pyandscripts/merge.py; code execution viaexec()inscripts/generate_all_offline.py. - Sanitization: Absent. Content is read and used directly without validation or escaping.
Audit Metadata