gate-dex-market
Fail
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [CREDENTIALS_UNSAFE]: Hardcoded default API and Secret keys are found in
references/openapi/_shared.md,references/openapi.md, andscripts/gate-api-call.py. Although these are documented as public defaults, the inclusion of hardcoded credentials in skill instructions and scripts is a high-risk security practice. - [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from external Gate DEX APIs, including market quotes, token info, and security audits, creating a surface for indirect prompt injection.
- Ingestion points: Market and token data responses retrieved from
api.gatemcp.aiandopenapi.gateweb3.cc. - Boundary markers: The instructions lack specific delimiters or instructions to prevent the agent from executing instructions potentially embedded in external data.
- Capability inventory: Execution of Python-based API helpers and shell scripts, along with file system access for credential storage.
- Sanitization: API results are interpolated into the agent context without sanitization or validation procedures.
- [EXTERNAL_DOWNLOADS]: The skill describes an auto-update feature that retrieves code and documentation from the vendor's official GitHub repository at
github.com/gate/gate-skills.
Recommendations
- AI detected serious security threats
Audit Metadata