derive-review-rubricslisted

Derive custom review-plugin judges, rubrics, and scoring criteria from evidence — existing skills, PR review feedback, and accumulated agent logs — by finding where agents needed correction or skills failed to activate, then translating those patterns into scoring dimensions and anchors. Hands the rubric design to create-review-plugin for scaffolding. Use when the user wants a custom skill reviewer or scoring rubric grounded in how their agents actually behave — to grade skills consistently, or build review criteria from past PR feedback and agent logs.
wagneripjr/skills · ★ 0 · Code & Development · score 72

Install: claude install-skill wagneripjr/skills

This skill grounds a custom reviewer plugin in real evidence. It gathers two evidence streams, translates recurring patterns into rubric dimensions with anchors drawn from the evidence itself, then hands the design to `create-review-plugin` for scaffolding and validation. It does not scaffold, copy schemas, or run `tessl review run` — that is `create-review-plugin`'s job. ## Input The user provides the scope to ground the rubric in: - The skills to review (a directory of SKILL.md files, or "the installed skills"). These are the artefacts the produced plugin will score. - A PR window for review feedback (a PR number or a time period like "the last 2 weeks"). - Optionally a log window and provider filter. ## Step 1 — Gather evidence ### Stream A — skills and PR feedback 1. Read each target SKILL.md for its intended behaviour (the `description` trigger and the workflow body). This is the baseline the rubric scores against. 2. Invoke `/find-optimizations` over the PR window, scoped to skill / SKILL.md PRs. Consume its findings — each has a **Type**, **Summary**, and **Evidence** (PR numbers and the specific review comments). The types come from find-optimizations' configurable list in `.tessl/memory/improvement-types.md` (default: `skill`, `rule`, `hook`, `test`, `refactor`, `verifier`); read it for the names in effect. Recurring `skill`-type feedback is the strongest rubric signal. ### Stream B — agent logs (optional) Run `tessl agent-logs view --json --since <ISO timest