Slv Benchmark

Run benchmark tests and connectivity checks for SLV endpoints using shredstream, grpc, or rpc with region-aware configuration and API key support.

This skill appears to do what it says (generate geyserbench configs, check endpoints, and use an ERPC API key) but review these before installing: - Confirm the source: the registry metadata shows limited provenance and the skill.json points at a GitHub repo; verify that repo and the author (ValidatorsDAO) are the expected maintainers. - Be aware the agent will look for and read ~/.slv/api.yml to obtain an ERPC API key. If that file contains other secrets or you don't want the agent reading files in your home directory, do not install or remove/relocate the key beforehand. - The skill will prefer to run a local benchmark binary (geyserbench) if present. Running a binary means arbitrary code execution on your host — ensure geyserbench is from a trusted source and you understand its behavior. - Note internal inconsistencies: registry metadata lists no required binaries while skill.json declares 'curl' and optional 'geyserbench'; the version in skill.json (0.9.962) differs from the registry version (0.13.15) and the skill's homepage was listed as none while skill.json points to a GitHub page. These mismatches could be benign (packaging sloppiness) but merit verification with the publisher. If you trust the publisher and are comfortable with the agent reading ~/.slv/api.yml and optionally executing geyserbench, the skill is functionally coherent. If not, ask the author to clarify the discrepancies and to provide a verified homepage/repository before proceeding.

Type: OpenClaw Skill Name: slv-benchmark Version: 0.13.15 The skill is designed for Solana endpoint benchmarking and connectivity testing. It automates the generation of configuration files and the execution of the 'geyserbench' utility. While it accesses a local configuration file (~/.slv/api.yml) to retrieve an API key, this behavior is limited to the tool's specific domain and is consistent with its stated purpose. There is no evidence of data exfiltration, malicious execution, or harmful prompt injection in SKILL.md or AGENT.md.