query-metrics

The SKILL.md is mostly clear and reviewable, though one transparency concern around a potential hardcoded credential weakens full pre-use reviewability and warrants a closer look. The skill keeps instruction and data boundaries reasonably separate and does not push the agent to treat external content as policy. Tested scenarios did not show material prompt-injection risk, but confidence is low: behavior analysis was not run, so downstream behavioral impact compared to a no-skill baseline remains unvalidated, and finite testing cannot guarantee safety in every case.