skills/testland/qa/grpcurl-cli/Gen Agent Trust Hub

grpcurl-cli

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the grpcurl binary from FullStory's official GitHub repository release page during CI setup (github.com/fullstorydev/grpcurl).- [COMMAND_EXECUTION]: Invokes the grpcurl binary via the shell to interact with gRPC services, perform discovery, and run health checks.- [PRIVILEGE_ESCALATION]: The CI configuration example uses sudo to move the downloaded binary into the system path at /usr/local/bin/ to enable global execution.- [INDIRECT_PROMPT_INJECTION]: The skill processes responses from external gRPC servers which may contain untrusted data.
  • Ingestion points: Output from grpcurl commands (JSON or plaintext responses).
  • Boundary markers: None explicitly defined in the prompt instructions.
  • Capability inventory: File system write access via -protoset-out and -proto-out-dir flags, and subprocess execution via grpcurl and jq.
  • Sanitization: Standard handling of CLI output; no specific AI-level sanitization is implemented.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 12:34 PM
Security Audit — agent-trust-hub — grpcurl-cli