Skills
Navigate
skills/goldsky-io/goldsky-agent/feeds/Gen Agent Trust Hub

feeds

Warn

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill reads the local Goldsky authentication token from the path ~/.goldsky/auth_token to authorize API key generation and retrieval.
  • [COMMAND_EXECUTION]: The skill uses curl to interact with the Goldsky API for key management and data retrieval.
  • [EXTERNAL_DOWNLOADS]: The skill fetches the OpenAPI specification from https://edge.goldsky.com/data/openapi.json for service configuration.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests wallet balance and transfer data from external endpoints, which represents a surface for indirect prompt injection. 1. Ingestion points: SKILL.md (wallet query endpoints). 2. Boundary markers: Absent for API responses. 3. Capability inventory: SKILL.md (curl execution). 4. Sanitization: Absent for external data processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 6, 2026, 12:09 AM