jeecg-portal
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions direct the agent to execute included Python scripts (
scripts/build_portal.pyandscripts/modify_component.py) to handle complex JSON construction and API interactions. These scripts utilize standard Python library modules (e.g., urllib, json, argparse) for their operations. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it retrieves data from external API endpoints and interpolates that data into configuration files (designJson) which are then saved back to the server.
- Ingestion points: Data is fetched from API paths such as
/eoa/cms/eoaCmsMenu/treeList,/joa/designform/designFormCommuse/roleDegisnList, and/eoa/portalapp/portalCustomRoute/listwithin the included scripts. - Boundary markers: The instructions and scripts do not implement specific delimiters or warnings to ignore instructions that might be embedded in the retrieved data (e.g., a malicious CMS title).
- Capability inventory: The skill has the capability to execute shell commands (running the local scripts) and perform network write operations (POST/DELETE/edit) to the user-provided API backend.
- Sanitization: The scripts process external text (like titles and names) directly into JSON structures without explicit validation or escaping, which could lead to malformed JSON or injection if the source data is compromised.
Audit Metadata