linkfox-lingxing-erp
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/response_io.pyexecutes internal skill components usingsubprocess.run. It employs list-based argument passing, which is a secure method that prevents shell injection vulnerabilities. - [EXTERNAL_DOWNLOADS]: The skill documentation specifies dependencies on well-known Python libraries like
requests,pycryptodome, andqrcode, which are to be retrieved from standard package registries. - [DATA_EXFILTRATION]: The skill facilitates data exchange between the local environment, the official Lingxing ERP API (
openapi.lingxing.com), and vendor-owned LinkFox domains. These operations are required for the skill's documented functionality and target trusted infrastructure. - [CREDENTIALS_UNSAFE]: A static application identifier is hardcoded in
scripts/onboarding.py. This identifier is used for routing authentication requests to the vendor's login service and does not represent a sensitive user-specific secret. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from the external Lingxing ERP API, creating a surface for indirect prompt injection. It lacks explicit sanitization or boundary markers for the ingested content, though the data is primarily handled as structured JSON.
Audit Metadata