near-jsonrpc-skill
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches a curated OpenRPC schema from the author's official GitHub repository at
https://raw.githubusercontent.com/holon-run/uxc/main/skills/near-jsonrpc-skill/references/near-public.openrpc.json.- [COMMAND_EXECUTION]: The skill utilizes theuxcCLI tool to link and execute JSON-RPC commands against NEAR blockchain providers as described inSKILL.mdandreferences/usage-patterns.md.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external NEAR RPC providers, creating a potential vector for indirect instructions. - Ingestion points: External data returned from blockchain RPC calls (e.g., account state, block details, validator sets) as described in
SKILL.md. - Boundary markers: Absent; the skill does not define specific delimiters for isolating or labeling the content of the external RPC responses.
- Capability inventory: Shell command execution via the linked
near-jsonrpc-cliand the baseuxctool as documented inSKILL.md. - Sanitization: Absent; no explicit escaping or validation of the RPC response content is provided before it enters the agent context.
Audit Metadata