Compare patched mandoc binary renders against the mandoc-md baseline and produce diff reports.
npx skills add https://github.com/idank/explainshell --skill eval-render
Please help me install this Agent Skill. Skill: eval-render Source: https://github.com/idank/explainshell/tree/main/.claude/skills/eval-render Command: npx skills add https://github.com/idank/explainshell --skill eval-render
Evaluate and compare the rendering of a candidate mandoc binary against the current mandoc-md, producing a verdict (merge/regression/defer) and a structured handoff prompt for regressions.
Run the eval with a candidate mandoc path and optional worktree to render, compare, and classify results.
High-intent search queries and answers about installing and using this skill.
You can validate a patched mandoc binary by comparing its rendered Markdown output against the current tools/mandoc-md baseline. The evaluation produces per-page diffs and a deterministic report classifying any regressions found.
A mandoc render regression evaluation is an end-to-end process that executes baseline and candidate renders, aggregates per-page diffs, and produces a verdict of merge, regression, or defer for the patched binary.
You generate a structured handoff prompt by running an end-to-end render evaluation against a candidate mandoc binary. The resulting diff report identifies regressions and produces actionable guidance for worktree reviewers.
Yes, you can run the evaluation with a candidate mandoc path and an optional worktree. It renders and compares the candidate output against the current mandoc-md baseline to classify results and guide validation workflows.
You should use an automated diff report for mandoc validation when you need deterministic regression risk assessment before merging a patched binary. It aggregates per-page differences to provide a clear merge, regression, or defer verdict.
idankOrchestrate mandoc -T markdown bug diagnosis, fix, and validation workflows.
OzeroHAXCompare current screenshots against baselines to detect UI discrepancies.
opendatahub-ioAnalyze evaluation results to identify alignment gaps between automated checks and human assessment.
niznik-devCreate custom inspect-ai evaluation tasks with experiment-guided and standalone modes.
