Installs: 0
Used in: 1 repos
Updated: 18h ago
$
npx ai-builder add skill jleechanorg/erInstalls to .claude/skills/er/
# /er — Alias for /evidence_review Runs `/evidence_review`, which now includes the `/es` evidence-standards check as Step 4 (after evidence review completes). **Usage**: `/er [subject or path]` ## Action Invoke `/evidence_review` with the same `$ARGUMENTS`. This is a thin alias — all logic (evidence review + evidence standards + synthesis) lives in `.claude/commands/evidence_review.md`. **Hard rule — Unit-only evidence is NOT ALLOWED.** If the only proof for a claim is unit tests (Layer 1, mocked/isolated), the verdict must be INSUFFICIENT and the user must be explicitly warned. Minimum: Layer 2 end-to-end integration tests (real callstack, mocks only at external API boundaries). LLM/external-service behavior claims require real-service evidence. **Exception:** unit-only proof IS acceptable for non-production changes (docs, tests, tooling/scripts) or for production changes under 100 delta lines of non-test code.
Quick Install
$
npx ai-builder add skill jleechanorg/erDetails
- Type
- skill
- Author
- jleechanorg
- Slug
- jleechanorg/er
- Created
- 18h ago