Stats
Actions
Tags
From sophie
Review a unit's reading against the Sophie evidence-based pedagogy rubric, separating the deterministic floor from quality-ceiling judgments.
How this command is triggered — by the user, by Claude, or both
Slash command
/sophie:review-chapter <unit-id>The summary Claude sees in its command listing — used to decide when to auto-load this command
Review the reading for unit **$1**. 1. Export the packet (read-only; prints to stdout): If `sophie` is not on `PATH`, use `pnpm exec sophie …` (consumer repo) or `node packages/cli/dist/bin.js …` (the Sophie monorepo). If it errors with "no build artifact," build the course first (`astro build`, or your project's `pnpm build`) so `dist/.sophie/` exists. 2. Follow the **sophie-quality-reviewer** skill on the resulting packet: echo the deterministic floor (`prior_findings`) verbatim, judge every rubric dimension against the `reviewed_artifact`, and return the `FLO...
Review the reading for unit $1.
Export the packet (read-only; prints to stdout):
sophie export-ai-context --task quality-review --target $1 --output markdown
If sophie is not on PATH, use pnpm exec sophie … (consumer repo) or
node packages/cli/dist/bin.js … (the Sophie monorepo). If it errors with
"no build artifact," build the course first (astro build, or your
project's pnpm build) so dist/.sophie/ exists.
Follow the sophie-quality-reviewer skill on the resulting packet:
echo the deterministic floor (prior_findings) verbatim, judge every
rubric dimension against the reviewed_artifact, and return the
FLOOR / CEILING / VERDICT report with severity-tagged findings.
Review only what the packet contains; do not edit files (HITL — the instructor acts on the report).
npx claudepluginhub drannarosen/sophie --plugin sophie