coinbase-openapi-skill
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and downloads a curated OpenAPI specification from the vendor's GitHub repository at
https://raw.githubusercontent.com/holon-run/uxc/main/skills/coinbase-openapi-skill/references/coinbase-advanced-trade.openapi.json. This file is used to configure theuxctool for API interactions. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external Coinbase API responses, creating a surface for indirect prompt injection if those responses were to contain adversarial content.
- Ingestion points: Responses from Coinbase Advanced Trade REST endpoints defined in the OpenAPI schema.
- Boundary markers: None explicitly used in the instructions, though the skill relies on structured JSON processing.
- Capability inventory: Execution of HTTP requests and management of credentials via the
uxcutility. - Sanitization: The instructions provide specific guardrails, advising the agent to use the JSON output envelope rather than text and to parse stable fields first to ensure predictable behavior.
- [COMMAND_EXECUTION]: The skill includes a validation script (
scripts/validate.sh) that uses standard shell utilities such asjqandrgto verify the presence and format of the skill's files.
Audit Metadata