Use when calling a gRPC service by hand or from a script: listing and describing services via reflection or proto files, invoking unary and streaming methods with JSON input, TLS and plaintext options, auth headers and exit codes for smoke tests and CI gates. Uses grpcurl.
89%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The workflow includes a GitHub Actions configuration that reads deployment status events and payloads, which can contain outsider-authored text injected into deployment payloads.
deployment payload (github.event.deployment.payload.host)
content-type · 2 sites
The GitHub Actions workflow reads github.event.deployment.payload.host from deployment_status events and uses it directly in grpcurl commands without validation, allowing outsider-controlled deployment payloads to inject arbitrary hosts.
SKILL.md
252
${{ github.event.deployment.payload.host }}:443
SKILL.md
261
${{ github.event.deployment.payload.host }}:443