Append advisory UI observations to review-status.json using curated screenshots.
npx skills add https://github.com/rootspec/rootspec --skill rs-review
Please help me install this Agent Skill. Skill: rs-review Source: https://github.com/rootspec/rootspec/tree/main/skills/rs-review Command: npx skills add https://github.com/rootspec/rootspec --skill rs-review
Provide an advisory visual assessment of the rendered UI using a small curated screenshot set. This advisory stage runs AFTER static review has already written authoritative findings and does not overwrite them.
Run rs-review after static review to append advisory findings to review-status.json using the curated screenshot set.
High-intent search queries and answers about installing and using this skill.
An advisory UI review appends human-like visual observations to the llmFindings key in review-status.json without overwriting the authoritative static findings already recorded.
Yes, the advisory UI review runs after static review and only writes to the llmFindings key, leaving existing authoritative static findings in review-status.json completely unaltered.
An AI-assisted visual review is needed when you want to surface advisory observations about UI rendering, accessibility, and visual consistency using curated screenshots after static analysis is complete.
Add visual observations by running an advisory review that reads the curated screenshots defined in llmInputs.screenshots and appends its visual findings exclusively to the llmFindings key in review-status.json.
Yes, advisory visual review works in review workflows where an existing review-status.json holds authoritative findings and llmInputs.screenshots define the curated visuals for generating advisory observations.
The best way to review rendered UI visuals without altering static test results is to run an advisory review that references seed and philosophy documents for context while writing only to the llmFindings key.
golemfoundationGenerate agent markdown files from council plans and research notes using templates.
miznizzzReveal internal reasoning and confidence levels for AI conclusions.
jakubciszakParse, deduplicate, and aggregate SARIF 2.1.0 findings across multiple runs.
kevinpbuckleyCapture Unreal Editor window, viewport, and asset screenshots via ScreenshotService API.
